- Top of page
- Patients and Methods
The natural history of lamivudine-resistant hepatitis B virus (HBV) infection in renal transplant recipients (RTx) is unclear, despite its increasing incidence. Twenty-nine HBsAg-positive RTx with rising HBV DNA received lamivudine therapy. The course of lamivudine-resistant HBV infection was studied prospectively. During 68.7 ± 12.5 months of follow-up, 14 (48.3%) patients developed lamivudine resistance, at 10–35 months (mean 16.9 ± 7.0). All showed mutant sequences at codons 552 and 528 of the YMDD motif, while 13 patients demonstrated wild-type sequence at codon 555. Lamivudine resistance was unrelated to patient demographics, HBeAg status/sero-conversion, or genotype. Following resistance, HBV DNA and alanine aminotransferase showed an initial increase followed by spontaneous gradual reduction. The subsequent peak HBV DNA was lower (1.26 ± 1.09 × 109 vs. 6.26 ± 12.23 × 109 copies/mL, p = 0.011), while that of alanine aminotransferase was higher (196 ± 117 vs. 77 ± 47 iμ/l, p = 0.005), compared with pretreatment levels. Post-resistance hepatitic flare occurred in 11 (78.6%) patients. This was transient in four (36.4%), but became chronic in six (54.5%) patients. Decompensation was noted in one patient during this flare, but all survived. We conclude that drug resistance is prevalent in lamivudine-treated RTx. Despite a lower ensuing peak viremia compared with baseline, hepatitic flare is common. While most patients have spontaneous resolution, a minority may develop potentially fatal decompensation during the preceding exacerbation.
- Top of page
- Patients and Methods
The outcome of renal allograft recipients (RTx) is adversely affected by hepatitis B virus (HBV) infection (1–9). The magnitude of this problem is especially profound in endemic areas when greater than 10% of the general population are HBsAg-positive. Before the advent of effective antiviral agents, chronic liver disease developed in greater than 80% of HBsAg-positive RTx, and 37–57% of mortality in these subjects was attributed to liver complications (3,4,7,8). Data from short-term studies have shown that lamivudine suppressed HBV replication and was well tolerated in kidney transplant recipients (10–12). We have demonstrated that lamivudine treatment in HBsAg-positive RTx with increasing HBV DNA markedly improved their survival (9). Notwithstanding the clinical efficacy of lamivudine, the frequent occurrence of relapse after stopping treatment has led to the common practice of continuing treatment indefinitely (13,14). In this context, viral breakthrough is an escalating problem, affecting 16–32% of patients within 1 year, and the incidence increases with the duration of therapy (11,12,15,16). Resistance to lamivudine is associated with mutations in codon 552 within the YMDD (tyrosine-methionine-aspartic acid-aspartic acid) motif of the HBV-reverse transcriptase/polymerase, with substitution of methionine for valine or isoleucine (17–19). Mutations at positions 528 (with replacement of leucine by methionine) and 555 (with replacement of valine, leucine, or methionine by isoleucine) can be associated findings in the development of resistance to lamivudine and/or famciclovir (20,21).
HBsAg-positive RTx demonstrate increased viral replication and increased severity of liver disease, consequent to the effect of immunosuppression (1–9,22–24). Among liver allograft recipients, the development of lamivudine resistance is followed by resurgence of viral replication and transaminase levels (25). Although fulminant exacerbation of liver disease is uncommon, 10–20% of liver recipients with lamivudine-resistant HBV may develop progressive liver failure. We have previously reported resistance rates of 7.4% and 37.0% in RTx after 1 or 2 years of lamivudine treatment, respectively (9). Despite the increasing prescription of lamivudine to organ transplant recipients, there is little data on the virologic and clinical course of infection by lamivudine-resistant HBV variants in RTx.
- Top of page
- Patients and Methods
Twenty-nine HBsAg-positive RTx satisfied the inclusion criteria and were started on lamivudine at 46.6 ± 43.1 months (range 1.0–175.8 months) after kidney transplantation (Table 1). Twelve of the 29 patients had lamivudine treatment commenced within 12 months of kidney transplantation, while none had acute rejection or bolus steroid treatment within 3 months before starting lamivudine. None of the patients had received lamivudine before kidney transplantation. All the patients were already on a maintenance dose or a tapering course of immunosuppressants when lamivudine was started, and there was no further change in their immunosuppressive regimen. Seven patients underwent liver biopsy before treatment, which showed chronic active hepatitis in all cases. All had serum creatinine less than 200 μmol/L, and lamivudine was given to all subjects at a uniform dose of 100 mg daily throughout the course of therapy.
Table 1. Characteristics of 29 HBsAg-positive renal allograft recipients treated with lamivudine
|Age (years)||47.1 ± 8.8 [31–68]|
|Post-transplant follow up (months)||105.7 ± 45.4 [37.9–201.7]|
|Baseline parameters (when lamivudine was started)|| |
|HBeAg status +/–||15/14|
|ALT (ULN)||2.02 ± 2.64 [0.26–7.11]|
|Bilirubin (umol/L)||24.4 ± 31.9 [7.0–165.0]|
|HBV DNA (copies/mL)||5.04 ± 10.25 × 109[2.58 × 107–4.59 × 1010]|
|Prednisolone dose (mg/D)||7.59 ± 1.95 [5–12.5]|
|Cyclosporine trough level (ug/L)||113.2 ± 21.9 [78–220]|
Fourteen (48.3%) patients developed drug resistance during 1292 patient-months of exposure to lamivudine. The duration of follow up was 56.7 ± 12.5 months after starting treatment. Lamivudine resistance occurred at 16.9 ± 7.0 months (range 10–35 months) after starting treatment, and 11 (78.6%) patients developed resistance within the first 18 months of treatment. The resistance rate was 3.6%, 46.0%, and 57.9% after 1, 2, and 3 years of treatment, respectively (Figure 1). Lamivudine was continued in all patients who developed drug resistance. Each patient was infected by a single HBV genotype, with half showing genotype B and the other half genotype C. The two genotypic groups showed similar levels of HBV DNA and ALT at baseline and after the emergence of lamivudine resistance (data not shown).
Figure 1. Graph showing the increasing prevalence of lamivudine resistance with the duration of lamivudine treatment in 29 HBsAg-positive renal allograft recipients.
Download figure to PowerPoint
Before the development of drug resistance all the treated subjects demonstrated effective suppression of HBV DNA to below the sensitivity of detection, with concomitant reduction of ALT from 2.02 ± 2.64 ULN at baseline to a trough of 0.51 ± 0.22 ULN after treatment (p = 0.002). Fifteen patients (eight in the mutant group and seven in the nonmutant group) were HBeAg-positive at commencement of lamivudine, and five (two in the mutant group and three in the nonmutant group) achieved seroconversion to become HBeAg-negative after treatment. Lamivudine resistance preceded the occurrence of HBeAg seroconversion in the two patients of the mutant group. Precore or core promoter variants were detected in five (62.5%) of the eight HBeAg-negative subjects in the lamivudine resistant group and in seven (63.6%) of the 11 HBeAg-negative subjects in the nonresistant group.
Results of the INNO-LiPA HBV DR line probe assay demonstrated the presence of methionine (mutant) at codon 528 in all 14 patients who developed lamivudine resistance. At codon 552, eight patients showed valine (mutant), while six showed isoleucine (mutant). At codon 555, one patient showed isoleucine (mutant), while 13 patients showed the wild-type amino acid valine. Three patients showed concomitant presence of wild-type virus and the lamivudine-resistant variant in the first blood sample that documented drug resistance. Among the 15 subjects without lamivudine resistance, four had HBV DNA levels less than 200 copies/mL (including one patient with HBeAg seroconversion) and thus were not assessable by the line-probe assay. Assay of the remaining 11 nonresistant subjects demonstrated infection by the wild-type virus. In patients infected by mutants, the same mutant pattern persisted during the course of follow up. The pattern of amino acid substitution at codon 552 was unrelated to HBeAg status (p = 0.627). Among the six HBeAg-positive patients, four showed valine while two showed isoleucine at position 552. The two patients with HBeAg seroconversion showed mutation patterns of L528M, M552V, and L528M, M552I, respectively, with valine detected at position 555 in both patients.
Predictors of lamivudine resistance
The gender, age, post-transplant duration, HBeAg status, e seroconversion rate (p = 0.608), time of starting lamivudine, and pretreatment levels of ALT and HBV DNA were similar between subjects who developed lamivudine resistance and those who maintained drug susceptibility (Table 2). Baseline ALT levels were abnormal in eight (57.1%) of the 14 patients who subsequently developed resistance, and in eight (53.3%) of the other 15 patients (p = 0.837). Body weight, height, body mass index, and cumulative steroid dose were also similar between the two groups (data not shown).
Table 2. Comparison of HBsAg-positive renal allograft recipients who did or did not develop lamivudine resistance
| ||Lamivudine-resistant HBV variants||p-value|
|Patients with (n = 14)||Patients without (n = 15)|
|Age (years)||46.8 ± 5.6||47.6 ± 11.2||0.913|
|Follow-up after transplantation (months)||103.4 ± 44.9||106.6 ± 46.4||0.913|
| ||[37.9–175.6]||[42.9–201.7]|| |
| when lamivudine started: +/−||8/6||7/8||0.573|
| latest: +/−||6/8||4/11||0.450|
|Months after kidney transplantation||40.8 ± 38.4||52.1 ± 47.8||0.662|
|when lamivudine was started||[1.0–103.7]||[3.0–175.8]|| |
|Peak ALT before lamivudine treatment (ULN)||1.45 ± 0.88||2.46 ± 2.08||0.238|
|Baseline ALT at starting lamivudine|
|treatment (ULN)||1.38 ± 0.83||2.42 ± 2.53||0.914|
|Peak HBV DNA before lamivudine|
|treatment (copies/mL)||6.26 ± 12.23 × 109||2.33 ± 3.41 × 109||0.172|
|Baseline HBV DNA at starting|
|lamivudine treatment (copies/mL)||6.14 ± 12.27 × 109||2.13 ± 10.25 × 109||0.311|
Outcomes of lamivudine resistance
HBV DNA and ALT levels both exhibited biphasic change following the emergence of lamivudine resistance. Each of these parameters showed an initial phase of increase followed by spontaneous gradual improvement at variable rates (Table 3; Figure 2). Circulating HBV DNA peaked at 16.2 ± 13.2 months after the emergence of drug resistance. This peak of HBV DNA (1.26 ± 1.09 × 109 copies/mL) was lower than both the pretreatment peak and the baseline HBV DNA level when treatment was started (p = 0.011 and 0.013, respectively). The latest HBV DNA level (1.49 ± 3.63 × 108 copies/mL) was significantly lower than its baseline when treatment was started (p = 0.002) and the postresistance peak (p = 0.001). In contrast, ALT reached its highest level at 20.2 ± 15.5 months after the emergence of drug resistance (p = 0.603 compared with the time for HBV DNA to peak after resistance developed). This ALT peak (3.70 ± 2.21 ULN) was higher than both its pretreatment maximum and its baseline (p = 0.005 and 0.002, respectively). The latest level of ALT (1.59 ± 1.51 ULN) was significantly lower than the postresistance peak (p = 0.003), but similar to the pretreatment baseline (p = 0.900).
Table 3. Serial ALT and HBV DNA levels in 14 HBsAg-positive renal allograft recipients who developed lamivudine resistance
|Patient no.||Duration of lamivudine treatment (months)||Months to development of lamivudine resistance||Baseline ALT (ULN)||Peak postresistance ALT (ULN)||Latest ALT (ULN)||Baseline HBV DNA (copies/mL)||Peak postresistance HBV DNA (copies/mL)||Latest HBV DNA (copies/mL)|
|1||59.8||13.0||0.91||7.21||5.74||2.50 × 109||1.02 × 109||6.79 × 106|
|2||53.9||14.0||0.57||8.04||1.89||9.90 × 108||2.40 × 109||1.39 × 109|
|3||30.9||13.0||1.38||2.74||1.62||6.14 × 109||2.12 × 108||2.11 × 108|
|4||64.9||15.0||0.92||1.60||0.55||3.17 × 109||7.79 × 108||1.87 × 107|
|5||67.9||14.0||2.74||2.23||3.43||4.59 × 1010||3.13 × 109||9.20 × 107|
|6||48.9||16.0||0.83||0.94||0.75||1.03 × 109||4.94 × 108||6.45 × 107|
|7||53.9||14.0||3.30||3.53||0.47||3.48 × 109||5.70 × 108||1.98 × 102|
|8||65.9||13.0||0.92||4.17||0.34||1.84 × 109||7.53 × 108||2.83 × 105|
|9||65.9||10.0||1.98||6.26||2.96||4.08 × 108||1.00 × 109||2.26 × 106|
|10||65.8||16.0||1.62||5.40||0.83||4.55 × 109||1.68 × 109||1.09 × 108|
|11||48.9||35.0||1.51||2.85||0.81||1.16 × 109||1.03 × 109||8.49 × 105|
|12||67.9||21.0||0.64||2.25||0.60||1.12 × 1010||3.74 × 109||3.85 × 107|
|13||32.9||13.0||1.45||2.87||1.11||3.11 × 109||4.70 × 108||5.66 × 105|
|14||65.9||29.0||1.49||1.77||1.19||5.39 × 108||2.90 × 108||1.51 × 108|
Figure 2. Serial HBV DNA and ALT levels in 14 HBsAg-positive renal allograft recipients who developed lamivudine resistance. The time-points included: baseline (when lamivudine was started), post-treatment trough, peak level after the emergence of lamivudine resistance, and latest follow up. Each box represents values from the first to the third quartile, and the whiskers indicate 1.5-fold the interquartile range greater or less than. The line across the box indicates the median.
Download figure to PowerPoint
Hepatitic flare occurred in 11 (78.6%) patients after the emergence of drug resistance. The abnormal biochemistry improved, concomitant with decreasing HBV DNA levels, within 6 months in four (36.4%) patients in the absence of therapeutic intervention or alteration of immunosuppressive medications. Six (54.5%) patients had persistence of the biochemical abnormality for more than 6 months. Most of the hepatitic flares were mild. Hyper-bilirubinemia was noted in two of the 11 patients, including one who had prolongation of prothrombin time. The latter was the only patient with severe flare of hepatitis after the emergence of drug resistance. This patient tested negative for HBeAg throughout the post-transplant course. Her favorable response to famciclovir was attributed to infection by an unusual variant that maintained famciclovir sensitivity despite lamivudine resistance (30). Famciclovir was successfully discontinued after 16 months of treatment. This patient remained stable under maintenance lamivudine over the subsequent 30 months, up to the time of writing. The latest ALT, albumin, bilirubin, and HBV DNA levels were 33 iμ/L, 40 g/L, 12 μmol/L, and < 200 copies/mL, respectively. None of the patients died. Hepatocellular carcinoma developed in two patients, both belonging to the nonresistant group.
- Top of page
- Patients and Methods
Nucleoside analogs target the DNA polymerase to suppress HBV replication. Owing to the absence of a proofreading mechanism for the DNA polymerase of HBV, drug-resistant HBV variants may emerge in the presence of selective pressure with prolonged nucleoside analog therapy (31). Resistance rates of 24%, 42%, 53%, and 70% have been observed in nonimmunosuppressed subjects after treatment with lamivudine for 1, 2, 3, or 4 years, respectively (32). The clinical impact of infection by lamivudine-resistant HBV variants is still being investigated. While it has been speculated that some YMDD variants may be less virulent in view of the reduced affinity for nucleotide substrates and the lower circulating levels compared with wild-type HBV (17,33), there is recent evidence that some variants demonstrate enhanced replication under the influence of lamivudine and can lead to severe hepatitis (34,35). To date, there is little information on the natural history of infection by lamivudine-resistant HBV variants in kidney transplant recipients.
Our results show that resistance to lamivudine affects greater than 40% of kidney allograft recipients within 2–3 years of treatment. Similar to nonimmunosuppressed subjects, drug resistance starts to emerge after about 9 months, and the prevalence increases with increasing duration of treatment. Although a lower rate for HBeAg seroconversion has previously been noted in patients with YMDD variants, our data show that resistance does not preclude HBeAg seroconversion (32). All the lamivudine-resistant HBV variants showed methionine (mutant) at codon 528, and all except one patient had valine (wild-type) at codon 555. There is no apparent association between HBeAg status and the type of amino acid substitution at codon 552. These preliminary results therefore do not suggest a link between replicative advantage and the pattern of amino acid substitution in the immunosuppressed patient. It is conceivable that the emergence of HBV variants is a dynamic process under the selective pressure of nucleoside analog treatment, as exemplified by the concomitant presence of drug-resistant and wild-type HBV in initial blood samples from three patients. The influence of immunosuppressive medications further confounds the complex virus–host interactions in organ transplant recipients. In the single patient who showed severe clinical deterioration after the emergence of YMDD variants, the initial favorable response to famciclovir was attributed to the preservation of wild-type sequence at position 528 (30,36). Follow-up blood samples in this patient showed substitution of leucine (wild-type) by methionine (mutant) at codon 528. Yet despite the apparent virologic resistance, the level of HBV DNA continued to decrease and the clinical improvement continued. The waning effect of immunosuppressive medications might explain the favorable clinical course. Our data showed persistence of the same mutant pattern in individual patients during follow up. In this regard, there is emerging evidence that distinct lamivudine-resistant variants can emergence sequentially, and in some instances the wild-type virus may re-emerge despite continuation of lamivudine treatment (35,37).
Compared with those who maintained drug efficacy, the development of lamivudine resistance has been associated with higher HBV DNA levels at baseline, both in nonimmunosuppressed subjects and in liver transplant recipients (32,37). In the present series, the apparently higher pretreatment HBV DNA level in patients who subsequently developed resistance did not reach statistical significance. A bigger sample size is also required to investigate the other factors that have been implicated in lamivudine resistance, such as a high body mass index and male gender. Our data nevertheless illustrate the difficulty to predict the development of drug resistance on an individual basis.
Earlier investigations have shown that following the emergence of lamivudine resistance, HBV DNA and ALT levels were lower than their pretreatment baselines and those in untreated subjects, and that adverse clinical complications resulting from drug-resistant HBV variants appeared uncommon except in those with poor liver reserve or in liver transplant recipients (30). The resurgence of HBV load appeared to be a critical factor leading to clinical exacerbation in organ transplant recipients (38). It is pertinent to note the biphasic change of HBV DNA and ALT in our patients who developed drug resistance. The initial phase of HBV DNA increase lasted for about a year after the appearance of drug-resistant variants. This was commonly associated with exacerbation of hepatitis, affecting 78.6% of patients. It is noteworthy that while the peak mutant HBV DNA level was lower than that at baseline, the reverse applied to ALT. Peak ALT level after emergence of drug resistance was more than twice that at baseline. One should therefore be alerted to the risk of potentially severe exacerbation of liver disease following the development of drug resistance in kidney transplant recipients. Subsequent to the initial phase of increasing HBV DNA and hepatitic relapse, both parameters abated spontaneously. Nevertheless, half of these patients showed persistently abnormal transaminase levels. The mechanisms leading to the spontaneous improvement are intriguing and remain to be investigated. Persistence of the same mutant patterns in follow-up blood samples did not suggest the emergence of less virulent variants, although the line probe assay may not be capable to distinguish subtle genomic differences. Dynamic interactions between the HBV variant, the waning effects of immunosuppression, and the host immune response could play a pivotal role.
HBV genotypes C and B predominate in our locality, in contrast to Caucasian communities in which genotype D or A is more prevalent. It remains controversial whether genotype C is associated with more severe liver disease, compared with genotype B (28,39). Our results did not implicate significant differences between B and C genotypes with regard to the levels of HBV DNA or ALT, before or after the development of lamivudine resistance. However, potential genotypic influence on long-term complications cannot be totally excluded in view of the relatively short follow up in this study.
We conclude that approximately half of HBsAg-positive RTx develop drug resistance with resurgence of HBV DNA after prolonged treatment with lamivudine. Exacerbation of hepatitis is common after the emergence of drug resistance. Decompensation can occur in a minority, while approximately half of these patients subsequently develop chronic hepatitis. Regarding the treatment of lamivudine-resistant HBV variants in renal transplant recipients, the nephrotoxic effect of adefovir dipivoxil presents a valid concern (40). In contrast, entecavir holds promise as an effective and safe alternative (41). In view of the high risk of lamivudine resistance with prolonged therapy, further investigations are warranted to examine the relative merits of continuous or intermittent treatment; the latter for a defined duration based on HBV DNA monitoring. Preliminary data suggest that discontinuation of treatment is feasible in carefully selected patients with careful surveillance for relapse (9).