Early virological suppression is associated with good maintained response to adefovir dipivoxil in lamivudine resistant chronic hepatitis B


Henry LY Chan, Department of Medicine and Therapeutics, 9/F Prince of Wales Hospital, 30-32 Ngan Shing Street, Shatin, Hong Kong.
E-mail: hlychan@cuhk.edu.hk


Aim To determine the factors affecting the virological response to adefovir dipivoxil (ADV) among patients with lamivudine resistant chronic hepatitis B.

Methods Chronic hepatitis B virus (HBV) infected patients, who had virological relapse to lamivudine, were switched to ADV monotherapy.

Results Twenty-six patients were treated by ADV for 23 (12–41) months. At baseline, the median log HBV DNA was 7.70 (4.88–9.01) copies/mL. Six (23%) and 8 (31%) of patients had HBV DNA suppressed to below 1000 copies/mL at month 12 and the last follow-up, respectively. On linear regression, patients who had higher HBV DNA at baseline and month 6 have higher HBV DNA at month 12. On Cox proportional hazard model, the hazard ratio for each log step increase in HBV DNA at baseline and month 6 for HBV DNA <1000 copies/mL at the last visit was 0.39 (P = 0.010) and 0.47 (P = 0.027), respectively. Alanine aminotransferase, HBV genotype, rtL80 M mutation and log HBsAg did not affect the HBV DNA response.

Conclusions The response of lamivudine-resistant patients to ADV is suboptimal. Treatment with ADV when HBV DNA is low, and rapid viral suppression at month 6 increases the chance of maintained viral suppression.


One major goal to treat chronic hepatitis B virus (HBV) infection is the suppression of HBV replication. Maintained suppression of HBV replication has been shown to reverse liver fibrosis and reduce liver-related complications.1, 2 Lamivudine has been the most popular anti-viral agent for many years. Its efficacy is hampered by the high incidence of drug resistance up to 70% in 4 years.3 Combination with peginterferon can reduce the occurrence of lamivudine resistance but it is associated with various adverse effects.4, 5 Treatment with new anti-viral agents including adefovir dipivoxil and entecavir is associated with less problem of drug resistance.6, 7 However, the high cost of the newer agents allows lamivudine to retain its role in selected patient groups particularly in developing countries.8 For instance, the maintained virological response of lamivudine among HBeAg-negative patients with severe acute exacerbation, which is usually associated with vigorous host immune clearance, is over 70% in 3 years.9

Adefovir dipivoxil is the first registered drug to treat lamivudine resistance. There is no cross resistance between lamivudine and adefovir dipivoxil.10 Early studies have demonstrated potent viral suppression of lamivudine resistant HBV by either switching to or adding adefovir dipivoxil.11 However, recent series in Korea suggested that adefovir resistant mutations tend to emerge more frequently and earlier in lamivudine resistant patients than treatment naïve patients.12, 13 Higher HBV DNA at the time of starting adefovir dipivoxil is probably associated poorer HBV DNA suppression in 2 years,14 but its effect on the risk of adefovir resistance is controversial.12, 13 Compensatory mutations to lamivudine resistance at amino acids 80 (rtL80I/V), 173 (rtV173L) and 180 (rtL180 M) of the HBV polymerase increase the replicative fitness of HBV but its impact on the response to adefovir is uncertain.10

In this study, we described the virological response to adefovir dipivoxil in a cohort of lamivudine resistant patients. Serum hepatitis B surface antigen (HBsAg) quantitation was used as a surrogate marker to covalently closed circular (ccc) DNA.15, 16 We also aimed to determine the impact of baseline factors including HBV DNA, HBV genotypes and compensatory mutations as well as early HBV DNA response at month 6 on the maintained virological suppression by adefovir dipivoxil.



Chronic hepatitis B patients who had virological relapse on lamivudine and treated by adefovir dipivoxil 10 mg daily for at least 1 year in Prince of Wales Hospital, Hong Kong were studied. Virological relapse was defined as HBV DNA elevation of over 1 log from the nadir during lamivudine treatment. Lamivudine was continued with adefovir dipivoxil for 1–2 months or until the alanine aminotransferase (ALT) level was normalized. All patients were maintained on adefovir dipivoxil monotherapy after cessation of lamivudine. Patients had regular monitoring of liver biochemistry and hepatitis e antigen (HBeAg) status every 3–4 monthly. HBeAg seroconversion was defined as loss of HBeAg and appearance of antibody to HBe (anti-HBe) for at least two occasions separated by 6 months. Residual serum samples were stored in −80 °C freezer for HBV DNA and HBsAg measurements. HBV DNA was determined at the time of starting adefovir dipivoxil (baseline), month 6, month 12 and at the last follow-up visit. HBsAg quantitation was measured at baseline, month 12 and the last follow-up visit. Genotypic resistance to lamivudine and adefovir dipivoxil was determined at the baseline and last follow-up samples.

Laboratory assays

HBsAg assays

HBsAg was quantified by Architect HBsAg QT (Abbott Diagnostic, Germany) according to the manufacturer’s instruction.17 The range of HBsAg detection in undiluted samples by the Architect assay extended from 0.05 to 250 IU/mL. Samples with HBsAg titer higher than 250 IU/mL were diluted to 1:500 to 1:1000 to bring the reading to the range of the calibration curve.

HBV DNA assay

HBV DNA was extracted by QIAGEN QIAamp DNA Mini Kit (QIAGEN Inc., Chatsworth, CA, USA) according to the instructions of the manufacturer. HBV DNA was quantified by TaqMan real-time polymerase chain reaction (PCR) assay as described previously.18, 19 This assay was standardized by serial dilution of EUROHEP genotype D HBV standard which contained 2.7 × 109 viral copies per mL. The range of HBV DNA detection was from 102 to 109 copies/mL with correlation coefficient of the standard curve routinely greater than 0.990.

Drug resistance assay and genotyping

To detect mutants at the polymerase gene, PCR amplification was performed by primers 2802–2824 (5′GCCTCATTTTGTGGGTCACCATA3′) and 996–977 (5′TCTCTGACATACTTTCCAAT3′) and PicoMaxx High Fidelity PCR master mix (Stratagene, La Jolla, CA, USA) at initial denaturation at 95°C for 2 min, followed by 40 cycles of amplification (95 °C for 60 s, 55 °C for 60 s, 72 °C for 90 s) and final extension at 72 °C for 7 min. PCR products were bi-directionally sequenced using the BigDye terminator version 1.1 cycle sequencing kit (Applied Biosystems, Foster, USA) on an ABI 3100 genetic analyzer (Applied Biosystems) by primers 300–318 (5′TTGGCCAAAATTCGCAGTC3′) and 997–977. HBV genotypes were determined in the baseline samples by comparing with the previous HBV sequences published in the GenBank.20

Statistical analysis

The primary aim of this study was to determine the residual HBV DNA levels at various time points of adefovir dipivoxil treatment among lamivudine resistant patients. As HBV DNA less than 1000 copies/mL during treatment was associated with reduced risk of drug resistance as reported in previous studies,21, 22 maintained virological suppression was defined as HBV DNA lowered than this level at the last follow-up visit.

Statistical tests were performed by SPSS (version 11.0, Chicago, IL, USA). Continuous variables were expressed as median (range). HBV DNA and HBsAg were logarithmically transformed for analysis. Spearman’s correlation coefficient was used for correlation of HBV DNA levels and HBsAg. Linear regression analysis was performed to assess the predictive factors for HBV DNA levels at month 12. Cox proportional hazard model was used to determine the predictive factors for HBV DNA less than 1000 copies/mL at the last follow-up visit. Multivariate analysis was not performed due to the small sample size. Statistical significance was taken as P value less than 0.05. All statistical tests were two-sided.


Baseline characteristics

Twenty-six chronic hepatitis B patients had virological relapse on lamivudine and were switched to adefovir dipivoxil treatment for 23 (12–41) months. The clinical characteristics of patients were shown in Table 1. Twelve patients had ALT <2 times the upper limit of laboratory normal (ULN), 7 patients had ALT 2–10 times ULN and 7 patients had ALT >10 times ULN. Two patients had serum bilirubin >3 times ULN. Four patients had HBV DNA less than 6 logs while all the remaining 22 patients had HBV DNA over 7 logs. Patients who had higher log HBV DNA tend to have higher log HBsAg (R = 0.63, P < 0.001).

Table 1.   Clinical characteristics of patients at the time of starting adefovir dipivoxil. Categorical data were expressed as n (%) and continuous data were expressed as median (range)
Age (years)44 (22–65)
Duration of lamivudine (months)43.3 (8.3–73.6)
Duration of overlap of lamivudine and adefovir (months)1.2 (0–12.7)
HBeAg positive16 (62%)
Log HBV DNA (copies/mL)7.70 (4.88–9.01)
Alanine aminotransferase (IU/L)167 (18–1543)
Log HBsAg (IU/mL)3.64 (1.87–5.27)
HBV genotype B:C9:17
rtL80I/V11 (42%)
rtV173L1 (4%)
rtL180M25 (96%)
rtA181V/T0 (0%)
rtM204I/V26 (100%)
rtN236T0 (0%)

Virological response

The median change of log HBV DNA was −2.86 (range −0.18 to −5.02) copies/mL at month 6, −3.50 (range −0.43 to −5.67) copies/mL at month 12 and −4.06 (range −6.64 to 0.44) copies/mL at the last follow-up. The HBV DNA levels at different time intervals of treatment are shown in Table 2. Six (23%) and 8 (31%) patients had HBV DNA suppressed to below 1000 copies/mL at month 12 and the last follow-up, respectively. Among the 17 patients who have adefovir treatment beyond 12 months, five patients had further reduction of HBV DNA (−1.91; range −1.12 to −2.16 logs) at 37 (range 24–39) months, 10 patients had less than 1 log change in HBV DNA levels followed for 27 (range 15–41) months, and two patients had increased HBV DNA (1.01 and 1.99 logs) at 26 and 33 months, respectively. Seven of sixteen (44%) HBeAg-positive patients had HBeAg seroconversion during adefovir dipivoxil treatment. Among the seven patients with HBeAg seroconversion on adefovir, three of them had history of transient HBeAg seroconversion during the previous lamivudine treatment but HBeAg was reverted back to positive before the commencement of adefovir. The log HBV DNA at the last follow-up of the HBeAg seroconvertors was 3.45 (range 2.00–6.15) and that of the non-seroconvertors was 4.33 (range 2.00–8.37) (P = 0.24). None of the HBeAg-negative patients had reversion of HBeAg. No patient had lost HBsAg.

Table 2.   Number (percent) of patients with different HBV DNA levels at different time intervals of adefovir dipivoxil treatment
Log HBV DNA (copies/mL)6 months12 monthsLast follow-up
  1. * One patient has undetectable HBV DNA at month 6 and month 12, which was also the last-follow-up visit.

  2. † Two patients who had undetectable HBV DNA at month 12 had detectable HBV DNA (log HBV DNA 2.83 copies/mL and 3.01 copies/mL, respectively) at the last follow-up (26 and 37 months, respectively).

  3. ‡ Three patients had delayed response with negative HBV DNA at months 34, 35 and 37, respectively.

Undetectable1 (4)*3 (12)*,†4 (15)*,‡
>2–31 (4)3 (12)4 (15)
>3–47 (27)8 (31)10 (39)
>4–54 (15)6 (23)3 (12)
>513 (50)6 (23)5 (19)

At the last follow-up, only 1 (4%) patient developed rtA181 V mutation. This patient has persistently positive HBeAg and his log HBV DNA was 8.64 copies/mL at baseline, 7.03 copies/mL at month 6, 6.38 copies/mL at month 12, and 8.37 copies/mL at month 33. No patient developed rtN236T mutation at the last follow-up visit. Among the nine patients who have lamivudine resistant mutations before adefovir treatment and positive sequencing results at the last follow-up, 5 (56%) patients had lost the signature mutations at both amino acids 180 (reverted back to L) and 204 (reverted back to M).

The median change of HBsAg was −0.30 (range −1.84 to 1.03) IU/mL at month 12 and −0.36 (−1.76 to 0.62) IU/mL at the last follow-up. At both month 12 and the last follow-up visit, 20 patients had reduction in HBsAg level and seven patients had elevation of HBsAg level. There was no correlation between the change of log HBV DNA and change of log HBsAg at month 12 (R = 0.20, P = 0.34) or at the last follow-up (R = 0.20, P = 0.33).

Biochemical response in relation to HBV DNA

Twenty-one (81%) patients had normalization of ALT within the initial 6 months. The median log HBV DNA of these 21 patients was 3.55 (range 2.00–8.37) copies/mL at the last follow-up visit. Two (7%) patients had slow decline in ALT and normalized ALT at month 12. One of these two slow biochemical responders had rapid HBV DNA suppression to 1.87 logs at month 6 and negative HBV DNA at month 12, but the other patient had slow HBV DNA suppression to 5.30 logs at month 6, 3.96 logs at month 12 and negative HBV DNA at month 37.

Three (11%) patients did not have gradual ALT normalization. Two of them had fluctuating ALT between normal and three times ULN throughout 16 and 29 months of adefovir dipivoxil treatment, respectively. The trough log HBV DNA of these two non-responders were 6.15 copies/mL and 3.68 copies/mL, respectively. The remaining patient had flare-up of ALT up to 549 IU/L at month 8 and the ALT slowly declined to normal at month 14 till month 22. This patient had persistently elevated HBV DNA of more than 6 logs during the course of adefovir treatment but she has no genotypic resistance to adefovir dipivoxil.

Factors associated with HBV DNA response

The factors associated with HBV DNA response at month 12 were shown in Table 3. Patients with positive HBeAg and higher HBV DNA at the time of starting adefovir tend to have higher HBV DNA levels (i.e., poorer virological response) at month 12. Patients who have higher HBV DNA at month 6 and those who had less reduction of HBV DNA at month 6 also tend to have higher HBV DNA at month 12. On Cox proportional hazard model, only HBV DNA levels at baseline and month 6 affected the HBV DNA level at the last follow-up (Table 3). Patients who had higher HBV DNA at baseline and month 6 tend to have a lower chance of achieving HBV DNA less than 1000 copies/mL at the last follow-up. HBV DNA <6 logs at baseline and <4 logs at month 6 were associated with approximately 50% chance of maintained suppression of HBV DNA (Table 4).

Table 3.   Factors affecting HBV DNA levels at month 12 (by linear regression analysis) and HBV DNA less than 1000 copies/mL at the last follow-up (by Cox proportional hazard model)
VariableMonth 12Last follow-up
Beta95% CIP valueHazard ratio95% CIP value
  1. CI = confidence interval.

Age−0.037−0.089, 0.0150.151.020.96, 1.090.55
Male gender0.051−1.47, 1.580.950.470.048, 4.710.52
Duration of lamivudine0.001−0.039, 0.0410.971.080.98, 1.180.11
Duration of overlap of lamivudine and adefovir0.082−0.077, 0.240.300.540.29, 1.0040.051
HBV genotype C−0.073−1.23, 1.080.901.670.32, 8.790.55
rtL80I/V−0.29−1.40, 0.820.591.820.39, 8.450.44
Positive HBeAg1.290.25, 2.330.0170.780.17, 3.680.75
Log HBV DNA at baseline0.490.015, 0.960.0440.390.19, 0.800.01
Alanine aminotransferase at baseline0.000−0.002, 0.0010.891.000.998, 1.0020.93
Log HBsAg at baseline0.41−0.26,, 1.030.057
Log HBV DNA at 6 month0.860.67, 1.06<0.0010.470.24, 0.920.027
Reduction of HBV DNA at 6 months0.640.28, 0.990.0010.900.44, 1.830.77
Table 4.   Relationship of log HBV DNA at baseline, month 6 and at last follow-up
 Log HBV DNA at last follow-up
Log HBV DNA at baseline
 ≤62/4 (50%)1/4 (25%)1/4 (25%)
 >6–85/14 (36%)5/14 (36%)2/14 (14%)2/14 (14%)
 >81/8 (13%)4/8 (50%)1/8 (13%)2/8 (25%)
Log HBV DNA at month 6
 ≤31/2 (50%)1/2 (50%)
 >3–43/7 (43%)4/7 (57%)
 >4–52/4 (50%)2/4 (50%)
 >52/13 (15%)3/13 (23%)3/13 (23%)5/13 (38%)

There was a trend that the shorter the duration of overlap of lamivudine and adefovir dipivoxil, the higher the chance of achieving log HBV DNA <3 copies/mL at the last follow-up (Table 3). However, this observation was probably biased by the study protocol as the duration of overlap would be longer if patients did not normalize their ALT within 2 months, and this relationship was not observed at month 12. Other factors including age, gender, duration of lamivudine treatment, HBV genotypes, polymerase mutations at amino acid 80 (rtL80I/V), ALT level and HBsAg level at baseline were not associated with HBV DNA response at month 12 or the last visit. Mutations rtV173L and rtL180 M were not analyzed due to the much skewed patient distribution (Table 1).


In this cohort of chronic hepatitis B patients who had virological relapse on lamivudine, adefovir dipivoxil could suppress the HBV DNA by 4.06 logs at a median follow-up of 24 months. However, only 30% of patients had HBV DNA below 1000 copies/mL and most patients still have significant viral replication. All patients who had abnormal ALT levels had inadequate suppression of HBV DNA (higher than 1000 copies/mL), but 1 log rise of HBV DNA preceding ALT elevation (i.e., phenotypic virological breakthrough) was uncommon. Patients who had lower HBV DNA at the time of starting adefovir dipivoxil and lower HBV DNA at month 6 maintained better viral suppression.

Adefovir dipivoxil has modest potency in HBV DNA suppression. Among treatment naïve HBeAg-positive patients, adefovir dipivoxil 10 mg daily for 48 weeks can suppress HBV DNA by 4.5 logs and approximately one quarter of patients achieved undetectable HBV DNA by PCR.23, 24 No patient developed genotypic resistance to adefovir in 48 weeks. However, among lamivudine resistant patients, the emergence of genotypic adefovir resistance ranged from 6% to 18% in the first year.12, 13 Previous preliminary analysis did not find any significant difference in the HBV DNA suppression and HBeAg seroconversion between lamivudine resistant patients and treatment naïve patients on adefovir dipivoxil.12 Although we did not have a control group, the 3.50 log step reduction of HBV DNA and 12% PCR negativity at month 12 seemed to be inferior to the previous reports among treatment naïve patients. The reduction in log HBsAg, which possibly reflected the reduction in cccDNA, was also lower in our patients (−0.30 IU/mL) as compared with that in another report among treatment naïve patients (−0.58 IU/mL) at the end of 1-year adefovir dipivoxil treatment.15 Although the HBeAg seroconversion rate was very high (44%) in this cohort, most of these the seroconvertors had history of transient HBeAg seroconversion during the previous lamivudine treatment and their HBV DNA levels at the last follow-up were not significantly lower than those who failed to achieve HBeAg seroconversion. As direct sequencing is much less sensitive than mass spectrometry to detect mutant virus when it is present as a minor proportion with the wild type virus, we believe the true incidence of genotypic resistance to adefovir dipivoxil should be higher than the 4% as reported in our patients.12, 13, 25

The majority of our patients have already developed phenotypic resistance to lamivudine with elevated ALT levels (up to 26 times ULN) and HBV DNA >6 logs when adefovir dipivoxil was started. The delayed commencement of adefovir dipivoxil when phenotypic resistance developed may partly explain the suboptimal HBV DNA suppression in our patients.14 In line with the previous report, we found that starting adefovir dipivoxil early when the HBV DNA was low tend to associate with better maintained virological suppression. Furthermore, rapid HBV DNA suppression at month 6 can also predict better virological response as in the case of lamivudine and telbivudine.21 In a pooled analysis among treatment naïve patients, HBV DNA at week 48 of adefovir dipivoxil treatment could predict the emergence of adefovir resistance in 4 years.22 We did not found baseline ALT level a factor that could predict virological response.

We did not have a rigid protocol on the overlapping of adefovir and lamivudine. Most patients have an overlapping of the two drugs for 1–2 months and lamivudine was usually stopped when ALT level returned to normal. We did not find any relationship between the duration of overlap and the virological response. However, recent studies suggested that adding adefovir dipivoxil to ongoing lamivudine treatment might offer better viral suppression than direct switching to adefovir dipivoxil monotherapy among lamivudine resistant patient with higher baseline viral load.26 In a large cohort of HBeAg-negative lamivudine resistant patients, combination of adefovir and lamivudine was associated with similar HBV DNA suppression but less virological breakthrough and less adefovir resistant mutations as compared with adefovir monotherapy.27

The mechanism of the suboptimal virological response and higher rate of drug resistance to adefovir dipivoxil among lamivudine resistant patients is uncertain. Although lamivudine resistant patients with higher baseline HBV DNA tend to respond poorer as shown by us and others,14 the level of HBV DNA alone should not be the sole explanation because treatment naïve patients with positive HBeAg and high HBV DNA can respond well to adefovir dipivoxil.20, 23 One alternative possibility is the development of adefovir resistant mutation during prolonged lamivudine therapy. Adefovir resistant mutation at A181 has been reported in the baseline samples of 24% of patients who did not develop negative PCR after a median of 18 months treatment.26 In our study, we could not detect any adefovir resistant mutation at amino acids 181 and 236 of the HBV polymerase in the baseline samples. Other compensatory mutations to lamivudine resistance including rtL80I/V, rtV173L and rtL180 M did not seem to affect the response to adefovir dipivoxil. Further efforts should be undertaken to investigate the mechanism of this phenomenon.

In conclusion, our results confirmed that virological suppression by adefovir dipivoxil is suboptimal in the majority of lamivudine resistant patients. Early treatment by adefovir dipivoxil when HBV DNA is low is important to increase the chance of maintained virological suppression. Monitoring for profound suppression of HBV DNA at month 6 can also assist in predicting better treatment response. Based on our findings, HBV DNA <6 logs at the time of starting adefovir dipivoxil treatment and <4 logs at month 6 have a better chance of maintained suppression of HBV DNA. Alternative treatment such as entecavir or tenofovir should be considered among patients who have poor initial response to adefovir to avoid the emergence of multi-drug resistant HBV mutants.28–30


Declaration of personal and funding interests: None.