Background
Chronic hepatitis B (CHB) virus infection is a serious global health problem with approximately 3.6% of the world’s population suffering from chronic hepatitis B infection [
1]. Hepatitis B virus (HBV) is transmitted vertically (from mother to child at birth) and horizontally (from person to person) [
2,
3]. Perinatal infection of infants is from HBeAg-positive mothers, which is common in southeast Asia, while HBeAg-positive mothers are not common in eastern Europe, Africa, and the Mediterranean basin [
4,
5]. In these areas, children are infected by close contact with HBsAg-positive individuals during early childhood, and more importantly, up to 90% of perinatal infections become chronic diseases [
6]. The positive rate of hepatitis B surface antigen in children varies by region. It is reported that the highest rate observed is 8.8% in Uganda [
2].
The current treatments for children with CHB are interferon-α and four nucleoside analogs: IFN-α initiated in children 12 months and older; LAM for children 3 years and older; adefovir and tenofovir initiated in children 12 years and older; and entecavir initiated in children 16 years of age [
7]. LAM is the first and primary antiviral drug currently officially approved for children with CHB under 12 years old. In addition, LAM is also less expensive than the other three NAs. Therefore, LAM is the first choice, particularly for children in developing countries, despite a low genetic barrier against the increase in resistance.
Because there are too few studies published in English investigating lamivudine treatment in children with hepatitis B infection, a recent meta-analysis of work examining the management of CHB in children only included one relevant study [
8]. Moreover, recent emerging clinical studies have not been consistent. To synthesize research regarding lamivudine treatment of hepatitis B in children around the world, not only the English databases but also the Chinese databases were searched for the present meta-analysis.
Materials and methods
Search strategy
Relevant studies were found by searching the English-language databases (EMBASE, PubMed, Web of Science, the Cochrane Library, and ClinicalTrials.gov) and Chinese-language databases [China National Knowledge Infrastructure (CNKI) and the Chinese BioMedical Literature Database (CBM)], using the following strategy:(((newborn* or neonat* or infant* or child* or adolescent* or paediatric* or pediatric*)) AND (lamivudine OR ‘2,3 dideoxy 3 thiacytidine’ OR ‘3TC’ OR epivir OR ‘lamivudine, (2S-cis)-isomer’ OR ‘BCH 189’ OR ‘GR109714X’)) AND (HBV OR hepatitis B). We included all cohort and randomized controlled trials (RCTs). The search was conducted in December 2018. The reference lists of all retrieved review articles were manually searched for potentially relevant articles missed by the intelligent retrieval system.
Selection criteria
The inclusive clinical trials had to fulfill the following criteria: (1) Study design: RCTs, with retrospective and prospective cohort study designs (each group sample size > 10); (2) Subjects: children or adolescents under the age of 18 with chronic hepatitis B; (3) Treatment strategy: including a LAM (100 mg/day) monotherapy group and a placebo or general treatment group as a control group; (4) Outcome: including virological responses, such as rates of HBV response, HBeAg loss and HBeAg seroconversion, or biochemical responses, such as rates of Alanine Transaminase (ALT) and Aspartate Transferase (AST) normalization. The exclusion criteria were as follows: (1) duplicated data; (2) coinfection with hepatitis A, C, D, or E viruses or human immunodeficiency virus; (3) Wilson’s disease, autoimmune hepatitis, primary biliary cirrhosis, hepatocellular carcinoma, etc.; (4) any report without sufficient information.
Outcome measures
End-points were defined before the initiation of the study. The primary efficacy end-point was the rate of HBV virological response that was by definition as the proportion of patients with undetectable serum HBV-DNA after treatment. The secondary efficacy end-points were as follows: HBeAg conversion rate, HBeAg loss rate; HBsAg loss rate; and biochemical response, defined as the normalization of ALT and AST. Safety end-points were height and weight changes after treatment.
Study quality assessment
The revised Jadad quality scale was used to evaluate the quality of all 8 RCTs included in the meta-analysis by examining the description of the randomization and blinding methods and the description of deviations and drop-outs. Out of the 8 RCTs, only one received a Jadad score of 6, and the Jadad scores were 3 for four studies and 2 for the remaining three studies. All 8 studies claimed to be RCTs, while only one study reported the randomization method. Only one of the studies was blinded. Four studies described study withdrawals and dropouts in detail. The Newcastle-Ottawa Scale (NOS) was used to evaluate the five included cohort studies based on several standards, including the selection of cohorts, comparability of cohorts, and assessment of the outcomes. Of the five cohort studies, only one received a NOS score of seven, and the four other studies received a score of six.
Using the same data collection table, the data for each included study were extracted independently and in duplicate by the two authors (Aoran Luo, and Xiaoyan Jiang). The data were extracted for (1) study characteristics (author, year of publication, geographic locale, study design, regimen, duration of follow up and sample size); (2) patient demographics (age, sex) and baseline characteristics (HBeAg-positive percentage, alanine aminotransferase levels, and serum HBV DNA levels); and (3) study outcomes after treatment. Any disagreement between the reviewers was resolved by a third party (Hong Ren).
Statistical analysis
All statistical analyses were carried out with Stata (version 12.0). For each included study, the dichotomous results were presented using the odds ratio (OR) with a 95% confidence interval (95% CI), while the continuous results were presented using a standardized mean difference (SMD) with a 95% confidence interval (95% CI). The statistical heterogeneity was assessed by using chi-square and I-square (I
2) tests with the significance level set at
p < 0.1. If significant heterogeneity was not present in the data, a fixed-effects model was adopted for analysis; otherwise, a random-effects model was adopted. In addition, a Galbraith plot and a sensitivity analysis were employed to explore sources of heterogeneity. Finally, funnel plots were constructed for eligible outcomes, together with Egger’s tests to examine the possible publication bias [
9]. All
P values were two-sided. Apart from Cochran’s Q-test, the significance level was set at
P < 0.05.
Discussion
Currently, the first-line medication for hepatitis B patients in children under the age of 12 is mainly interferon and lamivudine. Because interferon is injected subcutaneously, while lamivudine is administered orally, lamivudine is more convenient for the patient and therefore more widely used. However, at present, there is only one large-scale clinical trial of lamivudine, published in the New England Journal of Medicine [
10]. Moreover, a recent meta-analysis of studies examining the management of chronic hepatitis B viral infection in children only included this study [
8], but related experiments have been published continuously in recent years. It is necessary to conduct a meta-analysis to summarize the worldwide data. The present meta-analysis was performed by carefully reviewing 8 individual RCT studies and 5 cohort studies to compare outcomes related to chronic hepatitis B in children between LAM therapy and control groups. Subgroup analyses were primarily addressed through the study design or by the area in European and American countries and China.
The current meta-analysis demonstrates that lamivudine is effective in reducing the rates of HBV DNA response, HBeAg loss, HBeAg seroconversion and HBsAg loss and improving liver biochemical parameters in children who are infected with HBV. In addition, our results show that lamivudine therapy has no negative impact on children’s height and weight. It was unexcepted that subgroup analysis by area showed that the HBeAg loss rate was similar between the two groups for both European and American countries, but for studies from China, the HBeAg loss rate was higher in the LAM group than in the control group. The following reasons may explain this result: (1) genotype A is prevalent in European and American countries, while genotype B and genotype C are prevalent in China; (2) there are too few relevant studies in European and American countries to achieve statistical significance in differences between the LAM and control groups.
Although lamivudine can effectively inhibit the growth of the virus and improve liver inflammation, the use of lamivudine still has two limitations: (1) pre-existing covalently closed circular DNA in the liver cannot be eliminated by nucleoside analogs, including LAM, so relapse is frequent when therapy is withdrawn; (2) the emergence of resistant strains called YMDD (tyrosine, methionine, aspartate, aspartate) could lead to virological breakthrough followed by biochemical breakthrough. Therefore, LAM should be used carefully in the treatment of hepatitis B in children. Some studies found that the combination of interferon and lamivudine can improve these two aspects of lamivudine defects. Chan et al. found that a lower rate of LAM resistance emerged with a combination therapy of peg-IFN and LAM (21%) compared with LAM monotherapy (40%) [
23]. Selimoglu et al. showed the rate of breakthrough to be 23.4% in children treated with IFN-α and LAM combination therapy. Hence, the combination of interferon and lamivudine may be a good choice, especially for children under 12 years of age [
24]. Moreover, large prospective randomized trials examining the use of nucleoside analogs with higher genetic resistance barriers (Entecavir and Tenofovir disoproxil) in children with hepatitis B infection have been largely lacking. Otherwise, for patients under 12 years of age, there may be more effective treatment.
Some possible limitations of this study should be considered before generalizing our findings. First, 8 randomized controlled trials and 5 cohort studies were included, so not all of the included studies were randomized controlled trials. Second, of the 1556 patients included, 438 were from European and American countries, and the remaining 1118 were from China. Third, due to the limited number of studies included, information on individual patients was not detailed enough to evaluate the treatment effects in the different subgroups. Fourth, the durations of lamivudine use were different, which probably impacted the treatment effect of LAM therapy. Fifth, as the revised Jadad quality scale showed, the quality of the randomized controlled trials included here was not very high. Despite these limitations, the present meta-analysis summarized the newest worldwide data on lamivudine treatment for children. However, significant results were found primarily for Chinese but not for other populations. Large and well-designed studies in a variety of populations worldwide are needed to better generalize the present results.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.