Clinical course after stopping lamivudine in chronic hepatitis B patients with lamivudine-resistant mutants


Dr N. Leung, Department of Medicine, Alice Ho Miu Ling Nethersole Hospital, 11 Chuen On Road, Taipo, N.T., Hong Kong.


Background : The efficacy of lamivudine therapy in chronic hepatitis B is well established. However, drug-resistant YMDD mutants emerge with extended therapy. This may result in the resurgence of viral replication, the return of hepatitis and histological deterioration.

Aim : To study the safety of stopping lamivudine when the drug is no longer effective.

Methods : In the 5-year Asian Lamivudine Study, 34 patients from a single centre were included in this study. They had harboured YMDD mutants for at least 2 years. Lamivudine was discontinued and they were followed up at regular intervals. Clinical symptoms, liver biochemistry and viral serology were monitored.

Results : In a median follow-up of 20 months after stopping lamivudine (range, 7–39 months), 20 of the 34 patients (58.8%) had elevated alanine aminotransferase (ALT), 13 patients (38.2%) had elevated ALT one to five times the upper limit of normal and seven patients (20.6%) had an ALT flare (ALT more than five times the upper limit of normal with detectable hepatitis B virus DNA). There was no liver decompensation. ALT flare could be predicted by ALT over twice the upper limit of normal at the time of stopping lamivudine (P = 0.037).

Conclusions : It is relatively safe to stop lamivudine after YMDD mutants have emerged. ALT levels greater than or equal to twice the upper limit of normal at the time of stopping lamivudine have a higher risk for ALT flare.


Hepatitis B virus (HBV) infection is one of the most common and serious infectious diseases in the world. It is estimated that 400 million people world-wide are suffering from chronic hepatitis B infection.1 Chronic hepatitis B in Asia is mainly acquired at birth and during early childhood, leading to major mortality and morbidity in adult life. Approximately 25% of chronic hepatitis B patients die prematurely due to complications of liver cirrhosis or hepatocellular carcinoma.2–5 Lamivudine is the first oral therapy for the treatment of chronic hepatitis B. Although lamivudine is effective in viral suppression and has minimal safety issues,6–8 it is not able to eradicate the virus in a significant proportion of patients, and extended therapy runs the risk of the emergence of drug-resistant mutants. There is a clinical dilemma with regard to the benefit and safety of stopping lamivudine after drug-resistant mutants have emerged.

A placebo-controlled study in Asian patients with hepatitis B e antigen (HBeAg)-positive chronic hepatitis B (the Asian Lamivudine Study) has shown that lamivudine suppresses HBV replication, enhances HBeAg seroconversion and improves necroinflammation compared with placebo at the end of 1 year of treatment.8 Increased HBeAg seroconversion following extended lamivudine treatment has been reported.9, 10 However, lamivudine-resistant HBV mutants at the YMDD motif of the HBV polymerase genome (YMDD mutants) emerge at a cumulative rate of 24% after 1 year and 70% after 4 years of lamivudine treatment.8–11 During long-term therapy, the clinical benefits of lamivudine are greatest in patients free of YMDD mutants. A significant proportion of patients with YMDD mutants show a return of hepatitis and histological deterioration.10

Stopping lamivudine treatment before HBeAg seroconversion is often associated with a relapse of hepatitis. Even in patients who have achieved HBeAg seroconversion, 23–50% show a relapse of hepatitis on cessation of lamivudine,12, 13 and some show hepatic decompensation with hepatitis flare.14 Apart from organ transplantation patients, few publications have addressed the issue of stopping lamivudine treatment in patients harbouring YMDD mutants.15 The main concern is the possible risk of hepatitis flare when the wild-type HBV re-emerges.16

This study describes the clinical course of patients with YMDD mutants after stopping lamivudine therapy. Analysis was performed to identify the predictive factors of biochemical flare. This provides guidance for clinicians on the safety of stopping lamivudine therapy in patients with YMDD mutants in regions in which adefovir and entecavir are not viable therapeutic options.

Materials and methods


In the Asian Lamivudine Study, 358 patients with chronic hepatitis B were recruited from seven Asian centres in this double-blind, placebo-controlled, randomized trial of lamivudine therapy for 5 years.8 Briefly, eligible patients included males and females, 16–70 years of age, with detectable hepatitis B surface antigen (HBsAg) and HBeAg, serum HBV DNA levels of at least 5 pg/mL (as determined by solution hybridization assay; Abbott Diagnostics, Chicago, IL, USA) and alanine aminotransferase (ALT) levels less than 10 times the upper limit of normal (ULN). They were randomized to receive daily lamivudine, 25 mg or 100 mg, or placebo in the first 3 years. In the fourth and fifth years, all patients received open-labelled lamivudine 100 mg. Informed consent was obtained from each patient included in the study. The study protocol was approved by the Ethical Committee of our centre.

Fifty-eight patients from our centre participated and 56 completed the 5-year Asian Lamivudine Study. The emergence of YMDD mutants was detected at the end of each year by restriction fragment length polymorphism.17 They were detected in 44 patients in the 5-year period, with a cumulative prevalence at year 5 of 72.4%. One patient died of ALT flare and decompensation shortly after YMDD mutants were detected and another patient emigrated. Eight patients consented to a follow-on clinical trial of adefovir and lamivudine combination therapy after fulfilling the inclusion criteria of HBV DNA > 106 copies/mL plus ALT > 1.3 × ULN. Compared with patients who stopped lamivudine at 5 years, patients who were recruited into the adefovir study had higher HBV DNA levels (mean HBV DNA of 916 ± 1257 vs. 186 ± 250 MEq/mL, P < 0.001) and higher ALT levels (median ALT of 2.78 × ULN vs. 0.55 × ULN, P < 0.001). For the remaining 34 patients, lamivudine was stopped after completion of the 5-year study and they were followed up as reported in this study. All 34 patients had harboured YMDD mutants for at least 2 years. The demographic details of the patients are shown in Table 1.

Table 1.  Demographic details of the 58 patients in the original cohort before therapy and the 34 patients who harboured YMDD mutants when therapy was stopped at year 5
CharacteristicOriginal cohort (n = 58)Year 5 study cohort (n = 34)
  1. ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; ULN, upper limit of normal.

  2. Ranges of continuous variables are given in parentheses.

Male : female49 : 925 : 9
Median serum ALT (× ULN)1.93 (0.57–9.07)0.55 (0.31–4.41)
Median HBV DNA (MEq/mL)174.3 (1.8–822.6)161.0 (< 2.5–730)
HBeAg positive58 (100%)27 (79%)


After the cessation of lamivudine treatment at the end of year 5 in the Asian Lamivudine Study, patients were followed up at 4-weekly intervals for 24 weeks, and then at 3–6-monthly intervals as clinically indicated. Clinical symptoms, liver biochemistry and HBeAg/anti-HBe were monitored at each visit. Elevated ALT was defined as ALT above the ULN of 58 IU/L. ALT flare was defined as ALT ≥ 5 × ULN together with detectable HBV DNA. HBV DNA was checked at the end of treatment and during ALT flares.

Laboratory assays

HBsAg was tested using commercially available enzyme-linked immunosorbent assay kits (Abbott GmbH Diagnostika, Wiesbaden-Delkenheim, Germany). HBeAg and anti-HBe were measured by enzyme-linked immunosorbent assay (Sanofi Diagnostics, Pasteur, France). HBV DNA was measured by branched DNA assay (Chiron Diagnostic, Emeryville, CA, USA) and nucleic acid cross-linking assay (NAXCOR XLnt, NAXCOR, Menlo Park, CA, USA). These two assays have been shown to have similar sensitivity and range of detection from 105 to 109 viral copies/mL.18

Statistical analysis

Categorical data, including patient age and gender, HBeAg, HBV DNA and ALT status of the patients, were compared by two-tailed Pearson's chi-squared test or Fisher's exact test, as appropriate. Statistical significance was taken as P < 0.05.


Study population

All patients were of Chinese ethnic origin. Twenty-five patients (73.5%) were male and nine (26.5%) were female. The median age when lamivudine was stopped was 40 years (range, 26–56 years). The median ALT at the end of lamivudine therapy was 0.55 × ULN [range, (0.31–4.41) × ULN]. No patient had liver decompensation when lamivudine was stopped. The median serum total protein was 76 g/L (range, 68–87 g/L), albumin 39 g/L (39–43 g/L), globulin 38 g/L (29–46 g/L), bilirubin 9 µmol/L (2–43 µmol/L), prothrombin time 10.6 s (9.8–12.5 s) and International Normalized Ratio 1.06 (0.95–1.27). The median serum HBV DNA was 161 MEq/mL (range, < 2.5–730 MEq/mL). Liver biopsy was performed in all 34 patients at baseline and after 1 year, in 25 patients after 2 years and in 24 patients after 3 years of lamivudine treatment (Figure 1). At the end of year 3, seven patients (20.6%) had no fibrosis on histology, eight (23.5%) had stage 1 fibrosis, seven (20.6%) had stage 3 fibrosis and two (5.9%) had stage 4 disease/cirrhosis. They were followed up for a median of 20 months (range, 7–39 months) after lamivudine was stopped.

Figure 1.

Median necroinflammatory scores and fibrosis scores of the studied subjects at baseline and during lamivudine treatment. The number of patients in whom liver biopsy was performed is as follows: 34 at baseline, 34 at year 1, 25 at year 2 and 24 at year 3.

Biochemical profile

The biochemical profile of the patients after stopping lamivudine is shown in Figure 2. During post-therapy follow-up, 20 patients were detected with raised ALT: seven patients (20.6%) developed ALT flare and 13 patients (38.2%) had elevated ALT levels of > 1 × ULN and < 5 × ULN. Fourteen patients (41.2%) had persistently normal ALT levels.

Figure 2.

Median alanine aminotransferase (ALT) profiles. The three subgroups were classified according to the patients' ALT when lamivudine was stopped. ULN, upper limit of normal.

In the seven patients who developed ALT flare, the peak ALT value was (5.1–14) × ULN (Table 2). The peaks occurred at a median of 8 weeks (range, 4–64 weeks) after stopping treatment. None of these seven patients developed jaundice or hepatic decompensation during the flare. Two lost HBeAg. One had positive anti-HBe 37 weeks after the flare (41 weeks after stopping lamivudine) with normalization of the ALT level. The other patient had fluctuating ALT levels after an ALT flare at 12.7 × ULN and lamivudine was re-started at week 34. He eventually achieved HBeAg seroconversion with normal ALT 1 year later.

Table 2.  Characteristics of the seven patients with biochemical flare after stopping lamivudine
(week 0)
(week 24)
(week 0,
Peak (weeks after
stopping lamivudine)
  1. ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; UD, undetectable.

  2. The time when lamivudine was stopped was taken as week 0.

1++7302.2114.0 4
3+3504.4110.7 4
4++UD0.5 7.764
5++391.14 7.316
6++2690.64 5.258
7++3300.55 5.1 8

Of the 13 patients who showed elevated ALT of > 1 × ULN and < 5 × ULN after stopping lamivudine, all had fluctuating ALT levels between normal and a median peak of 1.53 × ULN [range, (1.16–4.17) × ULN]. None of these patients developed jaundice or hepatic decompensation. Only the patient with ALT peaking at 4.17 × ULN was re-started on lamivudine, and the hepatitis resolved.

Virological profile

Seven of the 34 patients with YMDD mutants had negative HBeAg and positive anti-HBe at the end of lamivudine treatment. Two patients reverted to positive HBeAg after stopping lamivudine. Two patients were negative for both HBeAg and anti-HBe, and one developed anti-HBe 5 months after lamivudine was stopped.

Twenty-seven patients had positive HBeAg when lamivudine was stopped. Only two had HBeAg seroconversion after stopping lamivudine. Their clinical course is described in the ‘Biochemical profile’ section.

Of the 10 patients with YMDD mutants during the 5-year period, but undetectable HBV DNA at the end of treatment, one had an ALT flare and six had mildly elevated ALT after stopping lamivudine. All seven patients showed a return of viraemia (median, 42.8 MEq/mL; range, 1.3–281 MEq/mL). In contrast, the three patients with persistently normal ALT did not show detectable HBV DNA throughout the follow-up period.

Analysis for predictive values of ALT flare

Analysis revealed that the only significant predictor for ALT flare was the ALT level at the time of stopping lamivudine therapy. Using ALT > 2 × ULN and ALT < 2 × ULN, two of two (100%) of the former and five of 32 (16%) of the latter had an ALT flare (P = 0.037). Gender, age, HBeAg status and HBV DNA level at the end of lamivudine treatment did not predict subsequent ALT flare after the cessation of lamivudine (Table 3). ALT and HBV DNA levels at baseline (before lamivudine treatment) were not included in the analysis because we did not believe that they would influence disease development 5 years later.

Table 3.  Univariate analysis of factors at the end of lamivudine treatment associated with the biochemical flare of hepatitis B after stopping lamivudine
FactorNumber of patients (n)Biochemical flare (n, %)P value
  1. ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; ULN, upper limit of normal.

 Male256 (24%)0.644
 Female 91 (11%) 
Age (years)
 ≤ 40163 (19%)1
 > 40184 (22%) 
 ≤ 2 × ULN325 (16%)0.037
 > 2 × ULN 22 (100%) 
 Positive276 (22%)1
 Negative 71 (14%) 
 > 106 copies/mL245 (21%)1
 ≤ 106 copies/mL102 (20%) 


Our data suggest that it is generally safe to stop lamivudine in patients harbouring YMDD mutants. Only one-fifth of patients developed significant biochemical flare and no patient developed hepatic decompensation after stopping lamivudine. Patients with elevated ALT at the end of lamivudine therapy tended to have a greater chance of biochemical flare, whereas HBeAg and HBV DNA status had no bearing on the clinical course.

Most (five of seven) ALT flares occurred within 6 months after stopping lamivudine therapy. Thus, close monitoring in the first 6 months post-treatment is essential, especially if ALT is elevated when lamivudine is stopped. As a proportion of patients still develop adverse clinical outcomes with YMDD mutants whether or not lamivudine is continued, it is important to consider the indications for lamivudine before commencing therapy.

Recently, Liaw et al. reported the clinical course after the cessation of lamivudine treatment in 128 patients with YMDD mutants in Taiwan.19 In their series, two-thirds of patients had ALT > 5 × ULN. These figures are higher than those in our report. In the Taiwan series, there were no pre-defined criteria for stopping lamivudine. It is probable that some of the patients may have had elevated ALT at the end of treatment. The different durations of lamivudine treatment and different durations of YMDD mutant emergence may also explain the different results. Moreover, eight of our 42 patients with YMDD mutants were recruited into another clinical protocol and were not included in this study. As all eight patients had elevated ALT levels and therefore a higher risk of developing a biochemical flare (based on the findings of the current study), the proportion of patients with biochemical flare might have been underestimated in this study.

One important finding of our study was that the ALT level at the time of stopping lamivudine is a predictor of biochemical flare afterwards. On stopping lamivudine, the replication of wild-type HBV inside the liver is unchecked and is usually followed by rebound viraemia.20 YMDD mutants are less replicatively competent and are naturally suppressed.21, 22 As elevated ALT is an indication of ongoing active immune clearance, it is logical that patients with elevated ALT will tend to have a more exaggerated immune response to the viral surge.

As HBeAg and HBV DNA status at the end of lamivudine treatment cannot predict biochemical flare afterwards, the viral load of YMDD mutants has no bearing on the immune response after the re-emergence of wild-type HBV. Whether the dynamic change of the ratio of mutants to wild-type HBV is pathophysiologically associated with hepatitis flare remains to be addressed in a separate study. The main aim of this study was to analyse the data in order to arrive at some practical interim suggestion for the management of the increasing number of patients with YMDD mutants.

One limitation of our study was the small number of patients involved. However, the patients in this study were unique as all had borne YMDD mutants for at least 2 years and all were started on lamivudine based on the same inclusion criteria of the Asian Lamivudine Study. Another limitation of this study was the lack of a control group that continued to take lamivudine in the presence of YMDD mutants. Nevertheless, five of the 34 patients (14.7%) who had YMDD mutants had ALT > 5 × ULN during the last year of lamivudine treatment (data not shown). This rate of biochemical flare is similar to, if not higher than, that after the cessation of lamivudine treatment. In addition, the baseline Knodell scores of this cohort were between 1 and 10. No patient had liver cirrhosis at baseline or on follow-up liver biopsies. Thus, the safety profile demonstrated in this study may not be generalized to patients with advanced liver disease and cirrhosis.

Adefovir dipivoxil is effective in suppressing lamivudine-resistant YMDD mutants and has been approved by the Food and Drug Administration.23–25 Entecavir is another promising agent under evaluation.26 Despite evidence of viral suppression and the low rate of resistant mutation associated with 1–2 years of adefovir therapy, data from large clinical trials are awaited. Potential nephrotoxicity on long-term therapy cannot be ruled out. The open-ended therapeutic strategy and advice against conception whilst on any nucleoside or nucleotide analogue create a dilemma for patients who have not completed their family. Cost becomes an additional issue when the cost-effectiveness of the treatment has not been defined.

In conclusion, stopping lamivudine in patients with YMDD mutants may result in ALT elevation, but with little risk of hepatic decompensation. This is a feasible option in regions in which adefovir and other anti-viral drugs are not readily available.


The Asian Lamivudine Study was supported by Glaxo Wellcome Research and Development UK, now GlaxoSmithkline.