Introduction
Despite the availability of infant vaccination programs and potent antiviral therapies, the prevalent infection of hepatitis B virus (HBV) has long been a health burden. Compared to the infection acquired in adulthood, HBV transmission via the mother-to-infant route or in early childhood leads to a higher risk of immune tolerance and chronicity of HBV [
1]. Theoretically, early intervention of chronic hepatitis B (CHB) can prevent disease progression to liver cirrhosis, hepatocellular carcinoma, liver failure or other end-stage events. However, due to the low coverage of therapies and the limited reference data for the treatment of children with CHB, the effectiveness of antiviral regimens has not yet been fully proven in paediatric patients [
2].
The variety of drugs adds complexity to the choice of therapy regimen. With the development of antiviral therapies over the past few decades, interferon alpha (IFNα) or pegylated interferon (peg-IFN) and high genetic barrier nucleos(t) ide analogues (NAs) such as entecavir (ETV), are currently recommended as first-line therapies in children with CHB [
3,
4]. The approved drugs for paediatric CHB patients are limited because of the strict licenced ages and insufficient clinical data on efficacy and safety. ETV is recommended for patients ≥2 years old and IFNα is recommended for patients ≥1 year old under close monitoring. With the approval of peg-IFN α-2a for CHB paediatric patients(≥3 years old) by the United States of America and the European Union in 2017, peg-IFN α-2a has also recently been recommended in paediatric CHB patients by international guidelines. Due to the less frequent injection and prolonged stability in vivo of peg-IFN compared to IFN-ɑ, peg-IFN is promising to replace the use of IFN-ɑ [
5]. ETV and peg-IFN, both of which are first-line drugs and are the most commonly used in paediatric patients, have distinctly different mechanisms in controlling the progression of CHB. ETV, one of the NAs competitively inhibiting HBV polymerases, is highly potent in viral suppression and is generally well-tolerated, but has resistance limitations and requires a prolonged treatment duration. Peg-IFN, an immunomodulatory agent, has a finite treatment duration, no resistance, and achieves higher hepatitis B e antigen (HBeAg) and hepatitis B surface antigen (HBsAg) serological response (SR) rates than NAs, but is not extensively used due to high costs and side effects [
6,
7]. A previous study compared the treatment response of continuous ETV monotherapy and finite peg-IFN α-2b initial therapy in CHB adults after a median of 92-week follow-up, and found that peg-IFN had relatively higher rates of HBeAg and HBsAg SR, while ETV had relatively higher rates of HBV DNA undetectability [
8]. The 2 drugs have unique non-overlapping advantages, and the optimal choice of initial treatment between ETV and peg-IFN for individual patients remains controversial, especially for children and adolescents with CHB, considering the limited clinical data in this population. In addition, the immune status and tolerability of drugs between paediatric and adult patients might be different; thus, referring to studies of adults might be inappropriate. Taking these controversies into consideration, more clinical data regarding the efficacy and safety of these 2 drugs for CHB children and adolescents are required. In this retrospective observational study, we comprehensively evaluated the treatment response and adverse effects in treatment- naïve CHB children and adolescents receiving peg-IFN α-2a or ETV as an initial antiviral therapeutic regimen in a 48-week follow-up.
Host immune factors, especially cytokine balance, play a key role in the immune response against pathogens and influence the disease outcome of HBV infection (clinical recovery or persistent infection). Previous studies have revealed a significant association between host immune cytokine profiles and disease progression or treatment response [
9,
10]. Cytokines can act not only as signalling molecules that initiate and regulate downstream immune activities by binding to the specific receptors of target cells, such as NK cells and cytotoxic T cells, but also, as effector molecules that directly affect viral replication [
11]. For example, IFN can promote antigen presentation to CD8+ T cells as well as phagocytosis in NK cells and monocyte–macrophages. It has been reported that the induction of serum IFN-γ levels during treatment is associated with virological control and HBeAg seroconversion [
12]. Cytokines play various roles in HBV infection, including cellular immune responses [e.g.,interleukin (IL)-2, interferon (IFN)-γ], humoral immune responses (e.g., IL-4, IL-6, IL-10), immunomodulation and immunosuppression [e.g., tumor growth factor-beta (TGF-β), IL-10], and mediating inflammation(e.g., IL-17, IL-22,IL-23) [
13]. To identify the potential immune signatures of responders to treatment, a series of cytokines related to CHB were chosen after a literature research, and an investigation of pretreatment cytokine profiles was performed in the study population.
Methods
Study population
CHB children and adolescents aged 2–18 years who attended the Children’s Hospital of Chongqing Medical University and started antiviral treatment from Jan 2013 to Oct 2020 were systematically reviewed with electronic medical records and laboratory inspection. Currently, ETV and peg-IFNα-2a are both first-line drugs for CHB. For patients up to the licenced age of the 2 drugs, the antiviral therapy (ETV or peg-IFN) was selected according to the patients and their parents’ choice. Doctors informed the patients and their parents of the route of administration of these 2 drugs, the duration of treatment, related side effects, indications and contraindications before selecting the antiviral therapy. Eligible patients were HBeAg-positive with elevated alanine aminotransferase (ALT) (> 1.5 times the upper limit of normal) and treatment-naïve at baseline, who received Peg-IFN α-2a (Pegasys, Roche) or ETV (Entecavir Dispersible Tablets, Chia Tai Tianqing) and completed the regular follow-ups for at least 48 weeks. The exclusion criteria were as follows: patients who received antiviral treatments other than Peg-IFN or ETV; patients with hepatic decompensation, hepatocellular carcinoma, liver failure, liver transplantation, chronic immunosuppression, coexistence of other liver diseases (e.g. autoimmune hepatitis, drug-induced liver injury or Wilson’s disease) or coinfection of other hepatotropic virus (hepatitis A, C, D, E, or human immunodeficiency virus); patients who did not completed the 48-week treatment or lost follow-up. In this retrospective observational study, a total of 70 CHB children and adolescents who met the inclusion criteria were included in the data analysis, consisting of 44 patients receiving ETV mono-therapy(0.03 mg/(kg*day)) and 26 patients initiating with Peg-IFN α-2a mono-therapy(104μg/(m2*week)). The study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of the Children’s Hospital of Chongqing Medical University (No:2019 []), and written informed consents were obtained from the children’s guardians.
Clinical parameter assessment
Patient case report forms were used to abstract medical records for demographic and clinical information. Information on their diagnosis and treatments was obtained by review of laboratory, pathological, radiological, and clinical records. Biochemical tests were performed using routine automated analyzers. Serum HBV DNA levels were measured by real time fluorescence quantitative polymerase chain reaction assays (Sansure Biotech,China), and the lowest limit of detection was 400 IU/mL. Serum HBsAg and HBeAg level were detected using commercial Chemiluminescence Microparticle Immuno Assay (CMIA) kits (Abbott GmbH & Co. KG, Wiesbaden, Germany). The lower limit of quantitative HBsAg determination was 0.05 IU/ml, and HBeAg-positive at values ≥1 S/CO (sample rate/cut off rate). The HBV genotypes were measured by real-time fluorescence quantitative polymerase chain reaction assays (Daan, China). HBV genotypes were determined by direct sequencing of the pre-S/S gene and comparison with the reference sequences in GenBank (NCBI). Grade of Inflammation and stage of fibrosis in liver histology were evaluated according to Scheuer’s criteria on a scale of 0–4.
Definition of outcomes
The evaluation of outcome was assessed at 48 weeks of antiviral therapy. The definition of outcomes was set according the international guideline for pediatric CHB [
3]. HBeAg /HBsAg SR was defined as loss of HBeAg/HBsAg with or without the emergence of hepatitis B e antibody (HBeAb)/hepatitis B surface antibody (HBsAb). HBeAg seroconversion was defined as loss of HBeAg/HBsAg with the emergence of HBeAb/HBsAb. Biochemical response was defined as normalization of elevated ALT levels. Virological response was defined as HBV DNA < 2000 IU/ml, and an undetectable HBV DNA level in serum was < 400 IU/ml. Primary non-response was defined as less than 2 log
10IU/ml decrease in HBV DNA levels from baseline after 24 weeks of therapy. Early virological response was defined as suppression of HBV DNA to undetectable (< 400 IU/mL) after 24 weeks of therapy. The safety of treatment was assessed by recording the adverse events at each visit as well as periodically monitoring the cardiogram, levels of blood ammonia and lactate for ETV group, and white blood cell counts, thyroid function and levels of autoantibodies for peg-IFN group. Virologic breakthrough was defined as more than 1 log
10IU/ml increase of HBV DNA levels during therapy.
Multiplex cytokine assay by quantibody® array kit
Serum samples were obtained from 26 of the included patients at baseline (on the day of therapy initiation) and stored at − 80 °C. A human chemokine quantibody® array kit (Raybiotech)-- a multiplexed sandwich ELISA-based quantitative array platform, was used per the manufacturer’s instructions. InnoScan 300 Microarray Scanner (Innopsys, arc d’Activités Activestre, 31,390 Carbonne-France) was used for fluorescence detection. The quantibody® array quantitatively measured the expression of a series of 51 chemokines (the list of cytokines was in Table S
1). After reviewing the medical records and clinical data, we divided the 26 patients into 2 outcome subgroups depending on whether they achieved HBeAg SR (
n = 13) or not (
n = 13) after 48 weeks of treatment.
Statistical analysis
Clinical data were analyzed with SPSS 26.0 (SPSS, Inc. Chicago, IL, United States). Normality test was applied to detect the distribution of data before further analyzing. Comparisons were performed using the Student’s t test or Mann-Whitney u test for continuous variables and the Chi-square test or the Fisher’s exact test for categorical variables, as appropriate. The association between HBeAg SR and baseline factors was assessed using univariate analysis and the binary logistic regression analysis. The classification of HBeAg SR and non-SR (NSR) using cytokine profiles was performed by the “Random Forest” package with R software (R Foundation for Statistical Computing, version 4.0.2). A fivefold cross-validation of random forest model was conducted to rank the contribution of each cytokine to discriminate HBeAg SR and NSR. Multidimensional scaling plot was used for visual comparison of cytokine levels between the 2 groups. The importance of individual cytokines was measured by Gini scores during the classification; cytokines with higher mean decreased Gini scores were considered to be more important than cytokines with lower numbers. Receiver operating characteristics (ROC) analysis was performed to evaluate the prediction value in HBeAg SR. All tests were 2-tailed, and P < 0.05 was considered as statistically significant.
Discussion
We are the first to conduct a comparison study on the efficacy and safety of peg-IFN and ETV therapy over a period of 48 weeks in children and adolescents with CHB. In regard to treatment response, ETV was the only independent favourable factor for both virological and biochemical response, which suggested its potent efficacy in inhibiting HBV DNA replication and improving liver inflammation. However, SR especially HBsAg SR, which is an important marker for a functional cure, was more frequently achieved in patients receiving peg-IFN than in those receiving ETV, which was in accordance with former studies in adults [
8,
14]. HBeAg or HBsAg SR are important treatment endpoints, as they are usually indications of immune control over HBV and increased survival in the long term. Previous studies have reported that HBeAg seroconversion rates could reach 25–36.3% and HBsAg seroconversion rates could reach 2.3–6% in patients receiving peg-IFN, while seroconversion rates were low (approximately 20% and 0–2% for HBeAg and HBsAg SR, respectively) and less durable in patients receiving ETV [
6,
15,
16]. Currently, few studies have performed head-to-head comparisons between the 2 drugs in CHB patients. In this retrospective comparison study in CHB children, we found that peg-IFN was associated with a better HBeAg or HBsAg SR than ETV in the multivariate analysis. The underlying mechanism of the better SR of peg-IFN may be due to its antiviral activity against intracellular HBV covalently closed circular DNA (cccDNA) [
17,
18]. Recent studies have revealed that IFN can not only induce cccDNA degradation by the concerted action of APOBEC-mediated deamination and ISG20-mediated nuclease digestion [
19], but also repress cccDNA transcription by epigenetic modification such as by modulating GCN5-mediated succinylation of histones in cccDNA minichromosomes [
20,
21]. However, HBV has also evolved an immune escape mechanism and can generate multiple factors, such as viral proteins, to disrupt IFN signalling pathways. The stable minichromosome cccDNA in the hepatocyte nucleus and the unique replication strategy of HBV lead to the generation of large amounts of HBV viral proteins, including HBsAg. High and persistent HBsAg levels in the circulating blood can impair the host HBV-specific immune response and contribute to the immune tolerance and chronicity of HBV infection [
22], while reduction of HBsAg could help recover the host anti-HBV immune response and predict the treatment response in the early phase [
23]. In our study, lower HBsAg levels at baseline were positively associated with both HBeAg and HBsAg SR, which similarly indicated the value of quantitative HBsAg in guiding antiviral therapy in CHB children and adolescents. Moreover, younger age(< 6 years) was also a favourable factor for both HBeAg and HBsAg SR. Studies have reported that children and adolescents, especially younger children, seemed to have higher rates of SR than adults [
24‐
26], which is possibly due to the relatively short incubation time of HBV and the less exhausted HBV-specific immune response of the host [
27,
28]. A higher level of ALT was an independent favourable factor for HBeAg SR, which was also in line with previous studies [
29,
30]. Subgroup analysis exploring the relationship between different ALT levels and HBeAg SR could be conducted in future studies with larger samples.
Even though children and adolescents have a relatively good probability of achieving the HBeAg and HBsAg SR, the total rates of response are still lower than desired. To optimize the current therapy, selecting potential responders before treatment and adapting the regimen during the early phase of treatment are necessary. Viral markers such as HBV genotype, quantitative HBV DNA, HBeAg and HBsAg have been widely explored to predict the treatment response, and the value of host immune factors in the prediction has also aroused attention recently. Previous studies have reported that the differentially expressed genes in liver tissues that associated with treatment response were mostly enriched in the cytokine-mediated immune response pathway [
31]. Impaired HBV-specific immune responses of NK cells and T cells and poor antiviral cytokine production contribute to the chronicity of HBV infection and might influence the treatment response [
28,
32]. To evaluate the association between host immune profiles and different clinical outcomes, we measured a panel of serum cytokines in 26 patients before treatment. We found that pretreatment cytokine profiles could significantly differentiate HBeAg SR from HBeAg NSR by a random forest model. Among the measured cytokines, IL-18 was the most important in the classification with the highest Gini scores and it remained significant in the ROC analysis. After adjusting for other important virological and clinical factors, the elevated serum level of IL-18 at baseline was still independently associated with HBeAg SR. IL-18, a proinflammatory cytokine, can stimulate IFN-γ release by NK cells, T cells, dendritic cells and B cells with inflammasome activation. Previous studies have proposed that IL-18 may have a close interaction with HBeAg and may be valuable in predicting HBeAg SR in anti-HBV therapy. It has been reported that HBeAg can suppress IL-18-mediated IFN-γ expression in NK cells via modulation of the NF-κB signalling pathway, which facilitates the persistent infection of HBV [
33]. In turn, elevated IL-18 can inhibit HBV replication in murine and hepatoma cell lines through induction of IFN-γ and other unrevealed mechanisms [
34,
35]. Clinical studies have also proven that the level of IL-18 was significantly elevated before HBeAg SR in HBV monoinfected and HBV-HIV coinfected patients receiving antiviral therapy [
36,
37]. Evaluating serum cytokines such as IL-18 may be a complementary tool to virological and clinical parameters and help identify potential responders and the optimal timing of anti-HBV treatment.
No serious adverse effects of the two drugs were observed in this retrospective study. It seemed that patients receiving peg-IFN had more frequent and various adverse effects in the early period of treatment. ALT flares tended to be more frequently observed in peg-IFN group than in the ETV group (11/26 vs. 4/44) and mostly occurred in the early phase of treatment ranging from week 3 to week 23(week 8.1 ± 1.3). Among the patients who had ALT flares during treatment, higher rates of HBeAg SR were observed in the peg-IFN group than in the ETV group. That is, 6 out of 11 patients in the peg-IFN group and 1 out of 4 in the ETV group achieved HBeAg SR, all of which occurred close to the time of the ALT flares, ranging from week 5 to week 27 (week 15.4 ± 3.1). This finding was in line with a recent observational study of ALT flares in adults with CHB, which detected that ALT flares during peg-IFN treatment were associated with the decline of HBsAg and HBV RNA and could predict subsequent HBsAg loss [
38]. Regarding virological breakthrough, it is worth noting that the occurrence of virological breakthrough was relatively frequent in the ETV group, mostly due to poor patient compliance. It is possible that patients, especially children receiving ETV, easily forget to take the drug, as it is taken daily. Therefore, patient education and practice should be further strengthened for the guardians of paediatric patients, especially for those who take ETV therapy, to improve the long-term prognosis.
Despite the positive results, there are a few limitations of this clinical study. First, the follow-up and efficacy evaluation of the 2 drugs were performed within 48 weeks and the outcomes were unclear for the long term. However, it might be reasonable to set the endpoints of a comparison study at a time when both drugs are not discontinued, because the recommended treatment courses of the 2 drugs are currently quite different. Peg-IFN is usually given as a finite course of 1 year, while ETV is usually given as a continuous course of at least 2 years [
16]. It remains controversial whether a more prolonged peg-IFN course might be more effective. Recent studies have reported that a prolonged treatment course (72–96 weeks) of peg-IFN had higher rates of HBsAg SR and sustained virological response in HBeAg-negative adults with genotype D or E [
39,
40]. In addition, although we detected that there were significant relationships between pretreatment cytokine profiles and treatment response, it is worth noting that host immune profiles were altered in a complex and dynamic manner. Monitoring the longitudinal changes in host immune profiles might be valuable for understanding how the host immune system reacts to antiviral therapies. Future prospective studies with larger sample sizes and longer follow-ups should be conducted.
In summary, we found that initial peg-IFN therapy was more effective in achieving HBsAg and HBeAg SR but not virological and biochemical responses compared to ETV during a 48-week follow-up in children and adolescents with CHB. These results further highlighted the role of IFN or peg-IFN in achieving a functional cure of CHB. We also found that significantly different cytokine profiles and IL-18 levels between patients with HBeAg SR and NSR. Individual variability in immune status and response to antiviral therapy suggests that more investigations are required to identify potential responders before treatment. Immunological markers such as elevated pretreatment IL-18 might supplement the viral markers and serve as promising predictors of the CHB treatment response.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.