SEARCH

SEARCH BY CITATION

Keywords:

  • Chronic hepatitis B;
  • entecavir;
  • renal function;
  • roadmap;
  • telbivudine

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author Contributions
  9. Transparency Declaration
  10. References

There are limited data comparing the clinical outcomes between telbivudine and entecavir. We consecutively enrolled 115 telbivudine-naive and 115 entecavir-naive chronic hepatitis B patients, who were matched for age, sex, hepatitis B e antigen (HBeAg) status and cirrhosis, and treated for at least 2 years or less than 2 years but had developed resistance. Except for the rate of HBeAg seroconversion, which was similar, patients in the entecavir group had better clinical outcomes than those in the telbivudine group for alanine aminotransferase normalization (85.2% vs 78.4%, p <0.048), undetectable HBV DNA (96.5% vs 74.8%, p <0.001), and viral resistance (0.9% vs 21.7%, p <0.001) after 2 years of treatment, After applying roadmap or super-responders concepts, entecavir still had better outcomes than telbivudine in undetectable HBV DNA and viral resistance. The cumulative incidence of hepatocellular carcinoma development was similar between telbivudine-naive and entecavir-naive patients (p 0.565). In renal function analysis, there were significantly more patients with estimated glomerular filtration rate (eGFR) category improvement in both the telbivudine and entecavir groups at year 1 (p 0.006 and p 0.047, respectively). The rate of virological improvement was significantly higher with entecavir than with telbivudine after 2 years of treatment, whether applying the concepts of roadmap or super-responders. The incidence of hepatocellular carcinoma was similar between telbivudine and entecavir. Both telbivudine and entecavir were associated with eGFR improvement, especially in patients with renal insufficiency.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author Contributions
  9. Transparency Declaration
  10. References

Although highly effective vaccines to prevent hepatitis B virus (HBV) infection have been available since 1982, there are still more than 350 million chronic carriers, 75% of whom reside in the Asia Pacific region [1]. These patients are at risk of developing hepatic decompensation, cirrhosis and hepatocellular carcinoma (HCC). With an expanded range of treatment options and a substantial improvement in the understanding of predictors of response to therapy, the management of patients with chronic hepatitis B (CHB) continues to evolve. Currently, there are five oral nucleos(t)ide analogues approved for the treatment of CHB [2], including three nucleoside analogues (lamivudine, telbivudine and entecavir) and two nucleotide analogues (adefovir and tenofovir). Each of these agents is effective in rapid and profound suppression of viral replication, facilitating hepatitis B e antigen (HBeAg) seroconversion, achieving alanine aminotransferase (ALT) normalization, and improving liver fibrosis. However, there are limited direct head-to-head trials comparing the different antiviral agents. The registration trials of telbivudine and entecavir were compared with lamivudine, and tenofovir was compared with adefovir [3-5]. Although our previous study compared the efficacy of telbivudine and entecavir, there are many limitations including the small size of patient groups, only 1 year of comparison, and a heterogeneous baseline [6].

The concept of roadmap, proposed by Keeffe et al. [7], uses the 24-week virological response to minimize long-term resistance. From the GLOBE trial, although the resistance rate at 2 years for telbivudine is 11%, application of a roadmap concept may reduce this rate. Further analysis from the GLOBE trial identified optimal baseline characteristics plus undetectable HBV DNA at week 24 after treatment (so-called super-responders) is associated with favourable outcomes after 2 years of telbivudine treatment. These were: (i) HBeAg-positive patients with baseline HBV DNA <109 copies/mL, ALT >2 × upper limit of normal (ULN) and undetectable HBV DNA at week 24; (ii) HBeAg-negative patients with baseline HBV DNA <107 copies/mL and undetectable serum HBV DNA at week 24 [8]. However, in real-world clinical practice, there are limited data to support this concept.

With higher numbers of patients now being treated for CHB, possible adverse events have gained more attention. One area of concern is renal function. Adefovir and tenofovir are both acyclic nucleotide analogues structurally, which have been shown to be nephrotoxic [9-12]. However, it should be noted that recent retrospective analyses from clinical studies demonstrated that long-term telbivudine treatment is associated with steady improvement in renal function, including in patients with pre-existing renal disease and those receiving tenofovir, although the potential mechanisms are unclear [13]. There is no published study indicating this finding and it is not known whether long-term treatment with entecavir affects renal function.

The present study aimed to compare the efficacy and safety of telbivudine and entecavir in patients who received therapy for 2 years.

Patients and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author Contributions
  9. Transparency Declaration
  10. References

Study design and patients

This was a retrospective single-centre match–control study. Between April 2007 and October 2012, a total of 115 CHB patients naive-treated with 600 mg telbivudine daily at the Chang Gung Memorial Hospital, Kaohsiung Medical centre for at least 2 years or less than 2 years and who had developed virological resistance were enrolled in this study. The sample population comprised 88 (77%) men and 27 (23%) women, with mean age (± SD) 52.9 ± 12.4 years. A total of 28 (24%) patients were positive for HBeAg and 57 (50%) had cirrhosis. To compare the efficacy and safety of telbivudine and entecavir, 115 hepatitis B surface antigen-positive patients treated with 0.5 mg entecavir daily for at least 2 years were selected randomly, who matched for age, gender, HBeAg status and cirrhosis. Of these 230 patients, 186 patents eligible for this trial were recruited from our previous study conducted in our centre (telbivudine, n = 94; entecavir, n = 92, respectively) [6]. The remaining 44 patients who were treated with telbivudine or entecavir between June 2007 and October 2010 and fitted the enrolled criteria were added to the study.

The therapeutic strategy for CHB patients was based on the criteria approved by the Bureau of National Health Insurance of Taiwan in 2008. Briefly, the criteria for treatment of CHB patients are as follows: (i) seropositivity for HBV surface antigen pulse decompensated liver disease; (ii) elevated ALT levels ≥5 × ULN (≥200 IU/L) for HBeAg-positive patients; (iii) elevated ALT levels between 2 × and 5 × ULN (80 < ALT <200 IU/L) with HBV DNA levels >105 copies/mL for HBeAg-positive patients without clinical evidence of cirrhosis; (iv) elevated ALT levels ≥2 × ULN (≥80 IU/L) with HBV DNA levels >104 copies/mL for HBeAg-negative patients without clinical evidence of cirrhosis; and (v) HBV DNA levels >104 copies/mL for patients with clinical evidence of cirrhosis. Clinical cirrhosis was defined by one of the followings: (i) ultrasonographic evidence of small liver with splenomegaly and/or (ii) presence of oesophageal or cardiac varices. Patients were excluded if they had any evidence of autoimmune hepatitis or markers of hepatitis C, hepatitis D and human immunodeficiency virus, or patients received chemotherapy or immunosuppressant agents, and significant intake of alcohol (20 g/day for women; 30 g/day for men).

Patients were followed up every 3 months or less for clinical assessment, including conventional liver biochemical tests, α-fetoprotein level, and serological hepatitis B markers (including HBeAg and antibody to HBeAg). Serial HBV DNA levels were assessed at baseline, and every 6 months after treatment. Virological breakthrough was defined as either an increase of serum HBV DNA of at least 1 log copies/mL from the nadir for patients with detectable viral load, or serum HBV DNA >100 copies/mL for patients with undetectable viral load during treatment [14]. The viral mutational analysis was determined using nested PCR and direct sequencing, as described previously [15], at the time of virological breakthrough. In addition, ultrasonography was performed for the surveillance of HCC every 3–6 months. If tumour was suspected, dynamic computed tomography or magnetic resonance imaging or liver biopsy studies were performed for confirmation. The diagnosis of HCC was based on the guidelines of the American Association for the Study of Liver Diseases [16]. All patients were treated by 15 experienced hepatologists at our institution, and were informed of the purposes of the study, and subsequently gave their written informed consent. Serum samples were collected from each patient at the time of their evaluation and frozen at −70°C until use. The study was approved by the Institutional Review Board of Chang Gung Memorial Hospital.

Efficacy assessment

Efficacy measures included the proportion of patients with ALT normalization (<1 × ULN), HBeAg seroconversion, undetectable serum HBV DNA (<60 copies/mL) by PCR assay, and virological resistance at 1 and 2 years, respectively. Sub-analyses of efficacy outcomes based on super-responder (for HBeAg-positive patients, baseline HBV DNA <109 copies/mL, pulse ALT level >2 × ULN; for HBeAg-negative patients, baseline HBV DNA <107 copies/mL) [8] and roadmap (undetectable HBV DNA at week 24) [7] concepts were also performed. Serum HBV DNA levels were quantified retrospectively, if there were no available data, using an Amplicor HBV monitor (Roche Diagnostic Systems, Branchburg, NJ, USA), from frozen serum samples, with a lower limit of quantification of 60 copies/mL. For results exceeding the upper detection limit (6.4 × 108 copies/mL), dilution was performed.

Furthermore, the incidence of HCC was analysed in patients excluding HCC development at baseline or within 1 year after treatment.

Safety assessment

The safety assessment aimed to measure the change of renal function during treatment. Renal function was assessed by the change of serum creatinine and estimated glomerular filtration rate (eGFR) from baseline to 1 and 2 years after treatment. The eGFR was calculated using the Modification of Diet in Renal Disease formula: eGFR = 186 × (creatinine) – 1.154 × (age) – 0.203 × 0.472 (if female) [17]. The change of eGFR from baseline by baseline eGFR categories (<60, 60–90, >90 mL/min/1.73 m2) was analysed, especially the percentage of patients shifting from 60–90 to >90 mL/min/1.73 m2 was assessed. Missing data were handled as last observation carried forward.

Data analysis

Data are presented as means ± SD, proportions, or median (range). Categorical variables were analysed by the chi-square or Fisher's exact test and Student's t test and Mann–Whitney U test were used to examine continuous variables with normal and skewed distribution, respectively. The renal function was compared using the Wilcoxon paired sample and Mann–Whitney U test for each group and between two groups, and McNemar test was used for eGFR shift. Covariates with p value <0.20 in univariate analysis were included in multivariate analysis using Cox proportional hazards models to determine whether pretreatment clinical and laboratory factors and on-treatment factors were significant in predicting HBeAg seroconversion and virological resistance. The cumulative rate of HBeAg seroconversion, virological resistance, and HCC and the differences in the curves were analysed using the Kaplan–Meier method and the log-rank test, respectively. All statistical analyses were performed using SPSS 17.0 for Windows (SPSS Inc., Chicago, IL, USA). Survival curves were constructed using SigmaPlot® 9.0. All p values were derived from two-tailed tests and a level of <0.05 was accepted as statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author Contributions
  9. Transparency Declaration
  10. References

Clinical characteristics at baseline

The baseline characteristics of the two study groups are shown in Table 1. The two groups are similar in terms of serum ALT levels (226.4. ± 371.6 vs 204.9 ± 294.7 IU/L) and HBV DNA levels (6.0 ± 1.6 vs 6.0 ± 1.3 log copies/mL). However, the proportion of subjects positive for HCC is greater in the telbivudine group (23.5% vs 13.0%, p = 0.041). The median duration of follow-up was 31 months (mean 30.2 months) and 34 months (mean 36.3 months) in the telbivudine and entecavir groups, respectively. In the telbivudine group, 115 (100%), 96 (83.5), 40 (34.8%) and 11 (9.6%) patients were followed up for 1, 2, 3 and 4 years, respectively; in the entecavir group, 115 (100%), 114 (99.1%), 56 (48.7%) and 24 (20.9%) patients were followed up for 1, 2, 3 and 4 years, respectively, at the time of writing.

Table 1. Baseline characteristics of study patients
 Telbivudine (n = 115)Entecavir (n = 115)p value
  1. Data were expressed as mean SD, no (%).

  2. ALT, alanine transaminase; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; Cr, creatinine; eGFR, estimated glomerular filtration rate; UNL, upper limit of normal.

Age (years)52.9 ± 12.453.6 ± 12.10.672
Gender (male: female)88:2788:271.000
HBeAg-positive (%)28 (24.3%)28 (24.3%)1.000
ALT (IU/L)226.8 ± 371.6204.9 ± 294.70.628
ALT >2 × ULN (%)68 (59.1%)69 (60.0%)0.893
ALT >5 × ULN (%)32 (27.8%)33 (28.7%)0.884
Liver cirrhosis (%)57 (49.6%)57 (49.6%)1.000
HCC (%)27 (23.5%)15 (13.0%)0.041
HBV DNA (log copies/mL)6.0 ± 1.66.0 ± 1.30.915
Creatinine (mg/dL)1.0 ± 0.51.0 ± 0.30.398
eGFR stage (%)  0.061
<60 mL/min1212 
60–90 mL/min5565 
>90 mL/min3719 
Diabetes mellitus (%)17 (14.8%)15 (13.0%)0.703
Hypertension (%)21 (18.3%)17 (14.8%)0.478

Comparisons of efficacy at year 2 after treatment

When all the patients receiving telbivudine (n = 115) or entecavir (n = 115) therapy were included in an intention-to-treat analysis of efficacy, the cumulative rates of ALT normalization, HBeAg seroconversion, HBV DNA undetectability and resistance between the telbivudine and entecavir groups are shown in Fig. 1. There are no significant differences between telbivudine and entecavir groups in HBeAg seroconversion at year 2 after treatment (46.4% vs 42.9%). The proportions of ALT normalization and undetectable HBV DNA are significantly greater in the entecavir group than the telbivudine group at year 2 after treatment (85.2% vs 78.4% p 0.048; 96.5% vs 74.8%, p <0.001). Twenty-seven patients developed the telbivudine-resistant mutation of M204I at different times during the study period: 9 (7.8%) patients at year 1; 16 (16.7%) patients at year 2 and 2 (5%) patients at year 3. Of 27 resistant patients, 2 developed the second mutation of L180M. The cumulative rates of resistance were 7.8%, 21.7% and 24.9% in the telbivudine group at years 1, 2 and 3, respectively, which was significantly greater than in the entecavir group (0%, 0.9% and 0.9% at years 1, 2 and 3, respectively, p <0.001). Only one patient developed entecavir-resistant mutations of L180M plus M204V and S202G at month 18 after treatment.

image

Figure 1. Comparison of efficacy results between telbivudine and entecavir at year 2 (the intension-to-treat population).

Download figure to PowerPoint

The outcomes of patients with favourable baseline (super-responders) proposed by Zeuzem et al. [8]. and on-treatment parameters (roadmap) by Keeffe et al. [7]. are shown in Fig. 2 and Fig. 3. For HBeAg seroconversion, there was a significant difference in the telbivudine group compared with the entecavir group only after applying the super-responders concept (83.3% vs 41.2% at year 2, p 0.008). Although the rate of virological resistance in the telbivudine group decreased from 21.7% to 11.8% and 13.5% at year 2 after applying super-responders and roadmap concepts, respectively, it was still significantly greater than in the entecavir group (p 0.003 and p <0.001, respectively).

image

Figure 2. Cumulative rate of hepatitis B virus e antigen (HBeAg) seroconversion between telbivudine and entecavir groups. Roadmap and super-responders concepts were proposed by Keeffe et al. [7] and Zeuzem et al. [8].

Download figure to PowerPoint

image

Figure 3. Cumulative rate of resistance between telbivudine and entecavir groups. Roadmap and super-responders were proposed by Keeffe et al. [7] and Zeuzem et al. [8].

Download figure to PowerPoint

The efficacy between the two groups according to HBeAg status is presented in Table 2. The entecavir group showed significantly greater DNA undetectability and lower resistance both in HBeAg-positive and HBeAg-negative patients after 2 years of treatment.

Table 2. Virological, biochemical, and serological responses after 2 years of therapy according to hepatitis B early antigen (HBeAg) serostatus (the intention-to treat population)
 HBeAg-positive patientsHBeAg-negative patients
Telbivudine (n = 28)Entecavir (n = 28)p valueTelbivudine (n = 87)Entecavir (n = 87)p value
  1. ALT, alanine aminotransferase; N/A, not applicable.

ALT normalization22 (78.6%)25 (89.3%)0.27564 (73.6%)73 (83.9%)0.095
HBeAg seroconversion13 (46.4%)12 (42.9%)0.788N/AN/AN/A
DNA undetectable17 (60.7%)26 (92.9%)0.00469 (79.3%)85 (97.7%)<0.001
Resistance9 (32.1%)0 (0%)0.00116 (18.4%)1 (1.1%)<0.001

To identify risk factors associated with virological resistance, Cox regression analysis with adjustment for baseline potential confounders (age, sex, HBeAg status, serum ALT and HBV DNA levels, and categories of liver disease), HBV DNA levels after 24 weeks of treatment, and antiviral agents showed that treatment with telbivudine (hazard ratio = 28.195, 95% CI = 3.822–208.006, p 0.001) and HBV DNA undetectable after 24-weeks treatment (hazard ratio = 0.286, 95% CI = 0.123–0.663, p 0.004) were significantly associated with virological resistance (Table 3). Furthermore, to identify which factors were independent predictors for HBeAg seroconversion, Cox regression analysis with adjustment for host, viral and treatment effects showed that only baseline ALT >200 IU/L (hazard ratio = 4.232, 95% CI = 1.705–10.506, p 0.002) was significantly associated with HBeAg seroconversion (Table 4).

Table 3. Univariate and multivariate analysis of factors predictive of virological resistance during telbivudine or entecavir treatment
 UnivariateMultivariate
VariableComparisonHR (95% CI)p valueHR (95% CI)p value
  1. HR, hazard ratio; CI, confidence interval.

Age (years)5>50 vs ≤501.463 (0.644–3.321)0.363  
SexMale vs female1.091 (0.442–2.690)0.899  
ALT (IU/L)>80 vs ≤800.880 (0.416–1.860)0.737  
HBeAgPositive vs negative1.833 (0.846–3.973)0.1251.038 (0.447–2.408)0.931
CirrhosisYes v. No0.875 (0.416–1.839)0.724  
Antiviral agentsLdT vs ETV29.776 (4.045–219.197)0.00128.195 (3.822–208.006)0.001

HBV DNA

(log copies/mL)

Per 1 unit increased1.438 (1.105–1.870)0.0071.171 (0.889–1.542)0.262
HBV DNA undetectable after 24 weeks of treatmentYes vs No0.220 (0.105–0.463)<0.0010.286 (0.123–0.663)0.004
Table 4. Univariate and multivariate analysis of factors predictive of hepatitis B e antigen (HBeAg) seroconversion during telbivudine or entecavir treatment
 UnivariateMultivariate
VariableComparisonHR (95% CI)p valueHR (95% CI)p value
  1. HR, hazard ratio; CI, confidence interval; S/Co, the ratio of the sample (S) to the cut-off (Co).

Age (years)≤50 vs. >502.116 (0.852–5.257)0.1061.890 (0.734–4.878)0.187
SexMale vs. Female1.057 (0.490–2.279)0.888  
ALT (IU/L)>200 vs. ≤2004.827 (2.027–11.491)<0.0014.260 (1.714–10.591)0.002
HBeAg titre (S/Co)≤50 vs. >502.429 (1.128–5.230)0.0231.826 (0.811–4.111)0.146
CirrhosisYes vs. No0.698 (0.295–1.655)0.415  
Antiviral agentsLdT vs. ETV1.259 (0.577–2.707)0.572  

HBV DNA

(log copies/mL)

Per 1 unit increased1.136 (0.867–1.486)0.355  
HBV DNA undetectable after 24-weeks of treatmentYes vs. No2.721 (1.030–7.188)0.0431.866 (0.653–5.335)0.245

Comparison of HCC development between telbivudine and entecavir

Of the 188 treated patients without HCC at baseline, HCC was diagnosed in 15 patients during treatment, five in the telbivudine group and ten in the entecavir group. Excluding five patients who developed HCC in the 1st year of treatment, 183 patients, 86 in the telbivudine group and 97 in the entecavir group, were enrolled for the Kaplan–Meier estimation (Fig. 4). The overall cumulative incidence of HCC development was 2.9% and 7.4% at years 2 and 3, respectively, i.e. 2.5% and 4.1%, respectively, in the telbivudine group and 3.1% and 7.5%, respectively, in the entecavir group (p 0.565; Fig. 4a). Further analysis found that there was no significant difference in HCC development between groups with and without virological resistance (p 0.315; Fig. 4c). However, the cumulative incidence of HCC development was significantly higher in patients with cirrhosis than in those patients without cirrhosis (6.9% and 17.7% at years 2 and 3 vs 0%, p <0.001; Fig. 4b).

image

Figure 4. Cumulative rate of hepatocellular carcinoma (HCC) development between (a) telbivudine and entecavir. (b) liver cirrhosis and non-liver cirrhosis. (c) resistance and non-resistance groups.

Download figure to PowerPoint

Estimated renal function assessment between telbivudine and entecavir

The change of creatinine and eGFR from baseline after 1 or 2 years of treatment in the telbivudine and entecavir groups are shown in Fig. 5. Both creatinine and eGFR significantly improved in the telbivudine and entecavir groups after 1 year of treatment (p 0.001 and p <0.001 in the telbivudine group and p 0.011 and p 0.026 in the entecavir group, respectively), but no significant difference was found at year 2. There was no significant difference in terms of the change of creatinine and eGFR at year 2 from baseline data between the telbivudine and entecavir groups. The eGFR shift, according to three categories (<60, 60–90 and >90 mL/min/1.73 m2), from baseline to years 1 and 2 are listed in Table 5. There were significantly more patients with eGFR category improvement in the telbivudine and entecavir groups at year 1 (p 0.006 and p 0.047, respectively), but there was no significant improvement at year 2 for either group. Subgroup analysis of eGFR mean change showed that both telbivudine-treated and entecavir-treated patients with worse baseline eGFR (<90 mL/min/1.73 m2) had better eGFR improvement (Fig. 6).

Table 5. Shift table of estimated glomerular filtration rate (eGFR; MDRD) from baseline to year 1 and 2 with telbivudine or entecovir therapy (McNemar test)
 BaselineMonth 12 (N)Month 24 (N)
eGFR (mL/min/1.73 m2)n (%)eGFR (mL/min/1.73 m2)p valuen (%)eGFR (mL/min/1.73 m2)p value
<6060–90>90<6060–90>90
LdT<60111820.00683320.303
60–9055037183841915
>903701027320923
Total103155477873140
ETV<60128400.04786200.067
60–9065542184122712
>901906138035
Total961352315783217
image

Figure 5. The mean of creatinine (a) and estimated glomerular filtration rate (eGFR). (b) between telbivudine and entecavir groups at baseline, year 1 and year 2.

Download figure to PowerPoint

image

Figure 6. Estimated glomerular filtration rate (eGFR) mean change from baseline to year 1 between telbivudine and entecavir groups

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author Contributions
  9. Transparency Declaration
  10. References

This match–control study offers a unique opportunity to evaluate the efficacy and safety of telbivudine and entecavir in CHB patients. Our previous study in which telbivudine-naive and entecavir-naive patients treated for 1 year were compared indicated that telbivudine-naive patients had higher rates of HBeAg seroconversion. Meanwhile by combining baseline ALT and 24-week HBV DNA, telbivudine-naive patients had the lowest virological resistance [6]. However, there are some limitations in that study, including the heterogeneous baseline of the study population, only 1 year of follow-up, and small population size. This time, except for extension of the study population and the duration of follow-up, the telbivudine and entecavir groups were matched by age, sex, HBeAg status and cirrhosis. To our knowledge, the present study is the first match–control study to compare the efficacy and safety of telbivudine and entecavir. With regard to efficacy, entecavir has a better response in ALT normalization, HBV DNA undetectability and viral resistance after 2 years of treatment. These results are similar to our previous study which only indicates the 1-year result [6]. Although a meta-analysis study indicated that entecavir and tenofovir are the most effective antiviral agents for CHB [18], there are many limitations in that study, including only having a comparison for 1-year treatment and variations in definitions, measurements, patient characteristics and protocols. In the present match–control, sing-centre study, most of these limitations are overcome. Hence, in the present match–control study, we confirm the poor efficacy of telbivudine compared with entecavir for CHB patients after 2 years of treatment.

It is well known that the concept of roadmap and super-responders, proposed by Keeffe et al. [7] and Zeuzum et al. [8], use the 24-week virological response and baseline HBV DNA plus ALT levels to individualize ongoing management in CHB patients to minimize long-term resistance and improve long-term efficacy. From the present study, we also compared the telbivudine and entecavir by the concepts of roadmap and super-responders. Although the 2-year cumulative resistance of telbivudine decreases from 21.7% to 13.5% and 11.8% after adjusting with roadmap and super-responders concepts, it is still significantly higher than those in the entecavir group (p <0.001 and p 0.003, respectively). In actual practice, besides the inherent problems of antiviral therapy, the cost of long-term drug use, which is associated with patient compliance, is also an important issue, especially in some Asia Pacific countries [19]. Therefore, some clinical physicians may choose telbivudine as the first antiviral agent because of its relatively low price in Asia, and then implement the concepts of super-responders and roadmap to result in low levels of resistance and may offer a high probability of HBeAg seroconversion. However, according to our analysis, despite applying super-responders and roadmap concepts in telbivudine-naive patients, there is still a significant difference in virological resistance compared with those treated with entecavir. The potential cost–effect benefit of using telbivudine as first-line agent might be negated by the development of drug resistance, which would necessitate the addition of adefovir, or switching to tenofovir.

Notably, the resistance rate seems much higher in the present study at 2 years after applying either roadmap or super-responders concepts compared with the GLOBE study. We could not explain what made the discrepancy between the present study and the GLOBE study. However, our data was from actual patients, in which the compliance might be poor compared with patients in clinical trials. Further studies using more data of this type are required to confirm this discrepancy.

Our study indicated the higher HBeAg seroconversion in the telbivudine and entecavir groups (46.4% and 42.9%) after 2 years of treatment. Further analysis found that only baseline ALT >200 IU/L is an independent predictive factor for HBeAg seroconversion. The results are compatible with the study by Liaw [20], which indicates that ALT >200 IU/L reflects a more robust immune clearance of HBV and induces a higher chance of HBeAg seroconversion during treatment. Another predictive factor is lower titre of HBeAg (<50 S/Co; the ratio of the sample (S) to the cut-off (Co)) before treatment (p 0.023 by univariate analysis), although it is not significant after multivariate analysis. In the present study, the mean HBeAg level was 294 S/Co (median (interquartile range): 139 (24.0–277.8)), and our population comprised 17 (30%) HBeAg <50 S/Co and 21 (37.5%) HBeAg <100 S/Co. The high probability of HBeAg seroconversion noted in the present study can be explained. Recent studies indicated that pretreatment HBeAg levels and the changing patterns of quantitative HBeAg resulted in prediction of an HBeAg response to lamivudine and entecavir [21, 22]. Further studies using quantitative HBeAg to predict a virological response in telbivudine are needed. To date, the information from head-to-head studies comparing HBeAg seroconversion between nucleoside analogues is limited. The present study first indicates that for treatment with telbivudine or entecavir, baseline ALT remains the strongest predictive factor for HBeAg seroconversion.

The difference in the rate of HBeAg seroconversion between the telbivudine and entecavir groups becomes smaller from year 1 to 2 (46.4% vs 35.7% at year 1 and 46.4% vs 42.9% at year 2). Although there was a higher incidence of HBeAg seroconversion in telbivudine-naive patients at year 1, no more improvement was noted at year 2. In contrast, the improvement in the entecavir group at year 2 shortens the difference in HBeAg seroconversion. From our previous study [6], mild improvement of HBeAg seroconversion from 6 to 12 months after telbivudine treatment was noted; in the present study, no more improvement is noted from 1 to 2 years. The HBeAg seroconversion reached a peak at the 1st year, which is different from the GLOBE trial's findings. Our case number is small (HBeAg-positive, n = 28) and the distribution of pretreatment HBeAg titre (S/Co) is wide, which might explain the high incidence of HBeAg seroconversion in the 1st year and no improvement in the 2nd year. Further study enrolling more HBeAg-positive patients is needed to provide the precise results.

Hepatocellular carcinoma is a potentially fatal consequence of HBV infection. Increasing evidence suggests that persistent presence of HBeAg and persistently high serum HBV DNA levels are associated with increased risk of HCC [23-25]. Antiviral therapy, by suppressing HBV replication and accelerating HBeAg seroconversion, should decrease the risk of HCC. To date, several studies have examined the preventive effect of nucleoside analogues. However, there are no studies directly comparing nucleoside analogues in reduction of the risk of HCC. The results presented here first clearly confirm that there was no significant difference in HCC development between telbivudine and entecavir-naive patients, even in patients developing virological resistance. Notably, no one developed HCC among the non-cirrhotic patients during 34 ± 9.9 (12–55; median 32) months of follow-up after the initial antiviral therapy. These data are consistent with those of previous studies. In a systematic review study that pooled 3881 treated patients, HCC occurred in 0.5% nucleos(t)ide analogue-naive patients without cirrhosis during a 46-month (range 32–108 months) period [26]. A meta-analysis by Sung et al., with 1267 treated patients (mostly given lamivudine), indicated that HCC was diagnosed in 1.8% of patients without cirrhosis [27]. In contrast to non-cirrhotic patients, cirrhotic patients were still at high risk of developing HCC (6.7% and 17.7% at years 2 and 3, respectively) after telbivudine or entecavir therapy. These data seemed to be higher than in the study by Liaw et al. [28]. Further analysis found that in the present study for the total 114 patients with cirrhosis, 65 (57%) had portal hypertension, as shown by splenomegaly or oesophageal or gastric varices, which is an independent factor for HCC development (p 0.036). In addition, the age in our populations is greater than in the GLOBE study (median, 54 vs 43 years), which might explain the difference. In our analysis, old age is weakly associated with HCC development (p 0.08). Therefore, clinical physicians should be very cautious that once CHB patients develop cirrhosis, strategies to treat as soon as possible are required to improve the outcomes of chronic HBV infection.

In recent years, renal function has been an important issue in patients treated with nucleos(t)ide analogues. Adefovir is dose limited by tubular toxicity [9], and tenofovir therapy is associated with increases in creatinine and kidney failure in human immunodeficiency virus/HBV co-infected patients[10-12]. In contrast, it is mysterious that data from a preclinical study for telbivudine indicate the improvement of eGFR after treatment [13]. However, until now, no real-world published study has proved this finding. From our analysis, not only telbivudine but also entecavir improves renal function after 1 year of treatment. Many factors affect renal function, including history of diabetes, hypertension and refractory ascites. Further subgroup analysis found that the improvement in renal function becomes apparent after excluding patients with a history of diabetes (= 32), hypertension (n = 38) and refractory ascites (n = 16), but the results of a predominant improvement in the 1st year and steady improvement during the 2nd year, remain. To date, the precise mechanism of these clinical findings is missing. The only thing we know is that nucleoside analogues seem to be beneficial for renal function compared with nucleotide analogues. Further in vitro and vivo study should be performed to identify these findings.

Overall, 28 patients developed drug resistance during the study period: 27 in the telbivudine group and one in the entecavir group. For patients who develop virological resistance to telbivudine, current guidelines recommend switching to or add-on tenofovir or add-on adefovir if tenofovir is not available [29-31]. Of 27 telbivudine-resistant patients, excluding two lost to follow-up and three who died due to HCC, 22 patients shifted to or added on other oral antiviral agents: three shifted to tenofovir; 15 added on adefovir; the other four shifted to entecavir. Of three patients changing to tenofovir, all developed HBV DNA undetectability; of 15 add-on adefovir, 12 had undetectable HBV DNA and the other three remain to have serum HBV DNA detected without any evidence of mutations; of four shifting to entecavir, only one became HBV DNA undetectable, two keep low viral loads (533 and 629 copies/mL, respectively), and one developed multi-drug resistance (A181T + M204I + T184I) at month 18. So far there is no published study indicating how to treat patients with telbivudine-resistance. From our follow-up analysis, comparing add on adefovir which remains 20% of serum HBV DNA detectable after 1-year rescue therapy, switching to tenofovir is better for patients with telbivudine resistance. Further prospective large-scale study is needed to identify this issue.

Treatment was generally well tolerated in telbivudine-naive and entecavir-naive patients. No serious adverse events were recorded during treatment. The incidence of myopathy, characterized by myalgia and elevation of creatine phosphokinase levels, was 0%. Although three patients suffered from myalgia in the telbivudine group, the serum creatine phosphokinase levels were all within the normal range. After receiving analgesia, the symptom improved without discontinuing telbivudine treatment.

In conclusion, except for the rate of HBeAg seroconversion being similar, patients in the entecavir group had better clinical outcomes than those in the telbivudine group after 2 years of treatment, After applying roadmap or super-responders concepts, entecavir still had better outcomes than telbivudine. In addition, the incidence of HCC is similar between telbivudine and entecavir after 2 years of treatment. Finally, we confirm that both telbivudine and entecavir have an improvement role in renal function, especially in patients with mild renal insufficiency. Further studies will help to understand the mechanism of this protective effect.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author Contributions
  9. Transparency Declaration
  10. References

This study was supported by grants CMRPG-8A0061G from Chang Gung Memorial Hospital, Taiwan.

Author Contributions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author Contributions
  9. Transparency Declaration
  10. References

The study was designed by T-H Hu. The first author, MCT, reviewed raw data, developed an outline of the manuscript, and reviewed and revised serial versions on the manuscript. The other authors were actively involved in the interpretation of the results, provided critical review of the manuscript and had the final decision on the article's content.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Author Contributions
  9. Transparency Declaration
  10. References