Comparison of renal safety of Telbivudine and Entecavir in Chronic Hepatitis B patients with renal impairment: A meta-analysis

Background:This study aimed to evaluate the renal safety of telbivudine(LdT) and entecavir(ETV) in chronic hepatitis B (CHB) patients with renal impairment. Methods:Studies published from January 1, 2010 to February 1, 2020 were identied using the PubMed, Web of Science, Scopus,Cochrane Library,ClinicalTrials.gov and CNKI (China National Knowledge Infrastructure).Finally,a total of 7 studies (1088 patients) with eGFR outcomes were retrieved and analyzed.The meta-analysis was conducted using RevMan 5.3. Results:The results of the 7 eligible studies analyzed suggested that the eGFR was both improved after LdT and ETV treatment.Compared with the baseline level,the eGFR was signicantly improved with LdT (7.02 mL/min/1.73 m 2 ) while slightly improved with ETV (1.72 mL/min/1.73 m 2 ) after 1 years of treatment. The eGFR was signicantly higher in the LdT therapy group than in the ETV group after 6 months(RR = 4.63, 95%CI: 0.73–8.54,Z = 2.33; P = 0.02) (cid:0) 1 year(RR = 3.35, 95%CI: 1.18–5.52,Z = 3.02; P = 0.002) (cid:0) and 2 years(RR = 11.00, 95%CI: 4.84–17.15,Z = 3.50; P = 0.0005) of treatment . Conclusion: Our meta-analysis of current evidence demonstrated that in CHB patients with impaired renal function, LdT could be the better choice than ETV.


Background
Hepatitis B virus (HBV) infection is a serious public health problem all over the world, approximately 257 million persons, or 3.5% of the population, were living with chronic HBV infection worldwide [1].It is also respected to its complications,including liver cirrhosis, hepatocellular carcinoma (HCC) and liverrelated mortality [2,3].
Treatment's primary targe in CHB is to prevent disease progression into serious complications by inhibiting of hepatitis B virus (HBV) DNA replication [4].To date,six nucleos(t)ide analogs(NAs) are used in CHB treatment,including three nucleoside analogs: lamivudine [LAM], telbivudine [LdT], and entecavir [ETV]; and three nucleotide analogs: adefovir dipivoxil [ADV],tenofovir disoproxil [TDF]and tenofovir alafenamide (TAF).These antiviral agents usually have relatively strong antiviral potency and low incidences of Resistance.However,they can also potentially cause serious adverse events, such as myopathy,neuropathy,lactic acidosis, and renal dysfunction [5].Among the adverse events, renal dysfunction is one of the risk factors,particularly in patients who already have renal impairment [6,7].
Chronic HBV infection may cause renal dysfunction through immune complex-mediated glomerular diseases.As renal excretion is the primary route of elimination of the mainstay of CHB therapy [7],nucleos(t)ide may also induce kidney toxicity through various mechanisms.,such as renal tubular injury, apoptosis, and mitochondrial toxicity [8].
Due to renal function is frequently impaired in patients with compensated CHB and patients usually require long-term treatment [2,8],When choosing an appropriate antiviral medicine,nephrotoxicity,which marked by a decrease in estimated glomerular ltration rate(eGFR),should be particularly considered [9].Among NAs,Telbivudine (LdT) and Entecavir (ETV) are regarded as rst-line antiviral agent in CHB patients with renal impairment.
Telbivudine and entecavir are used widely to inhibit hepatitis B virus (HBV) replication.However, Data comparing renal safety of these two antiviral agents in CHB patients with renal impairment are limited and lack systematic evidence base.Therefore,the main objective of this study is collate all available evidence and summarize the data to assess the renal safety of the nucleoside analogs LdT and ETV in this speci c population.

Search strategy
PubMed, Web of Science, Scopus,Cochrane Library,ClinicalTrials.gov and CNKI (China National Knowledge Infrastructure) were searched without language restriction to identify relevant articles published from January 1, 2010 to February 1, 2020. The search was performed with the following keywords:'chronic hepatitis B','telbivudine', 'entecavir','nucleos(t)ide Analogues','renal Function''estimated glomerular ltration rate',and their synonyms and related terms.

Selection Criteria
The following inclusion criteria were used to select studies for review: 1. Randomized controlled trials (RCTs), retrospective and prospective cohort studies.All studies had proper clinical information.

Study populations involving patients with chronic hepatitis B and renal impairment [eGFR between
30 and 90 mL/min/1.73 m2],and the results contained at least eGFR outcomes.
3. The study including a LdT group or ETV group.
The exclusion criteria were as follows: 1. The patients were coinfected with either hepatitis A, C, D, or E virus or with human immunode ciency virus (HIV)..

2.
The study without su cient information.
3. The study interventions did not include either LdT or ETV.

E cacy endpoints
The e cacy end-point was the change in estimated glomerular ltration rate (eGFR) from baseline,which was calculated by the Modi cation of Diet in Renal Disease (MDRD) [10].Most studies reported the renal function outcomes at the time range from 6 month to 2 years follow-up,so the study e cacy endpoints were analyzed at a time point of 6 months,1 year and 2 years.

Data collection
Two authors independently searched literature and extracted data using a predesigned data collection template, and discrepancies were discussed via discussion.The data were extracted for: (1) study characteristics:study design,year of publication,region, study type,Interventions,sample size and follow-up period.
(2) patient characteristics (age, gender) and baseline estimated glomerular ltration rate (eGFR);(3) study outcomes after treatment.The Inclusion and exclusion criteria were uniformly applied across all the publications.

Risk of bias
The Cochrane Collaboration Risk of Bias Tool was used to assess the risk of bias [11]. This tool comprised of seven criterias:(1)Random sequence generation;(2)Allocation concealment;(3)Blinding of participants and personnel;(4)Blinding of outcome assessment,(5)Incomplete outcome data;(6)Selective reporting;(7)other bias.
Three levels were used to assess the methodological quality:(1)low risk of bias;(2)high risk of bias; (3)unclear risk of bias.Assessment was independently performed by two authors, and disagreements were resolved via discussion.

Data analyses
All statistical analysis was conducted using Review Manager version 5.3 (The Cochrane Collaboration, Oxford, UK).For each eligible study, the dichotomous data were presented using the relative risk(RR) with a 95% con dence interval (95% CI),while continuous data were presented using the weighted mean difference (WMD).
The statistical heterogeneity between studies was assessed by using the chi-square (χ 2 ) and I-square (I 2 ) tests.,with signi cance set at P < 0.05,When I 2 > 50%, P < 0.05 was considered to be statistically signi cant, and the random effects model was adopted for meta-analysis; otherwise,the xed effects model was adopted [11,12]

Literature search
The study selection process is summarized in Figure 1.We identi ed 6102 relevant studies though database searches,After the initial screening and eligibility assessment phase,1972 redundant publications were excluded and, after referring to the titles and abstracts,a further 41 studies that not ful lling the inclusion criteria were rejected. The remaining 4 studies without su cient data were rejected, Finally,7 studies were enrolled in this meta-analysis.Five of the articles were retrospective cohort analyse [13][14][15][16][17] and two was an RCTs [18,19].A total of 1088 patients, 522 of whom were treated with LdT and 566 with ETV were included in this meta-analysis.The LdT dose used in the studies was 600 mg/day and that of ETV was 0.5 mg/day. Table 1 summarizes the basic characteristics of the included studies and patients .Three of the included studies were from mainland, China [13,16,19].{Liu, 2019 #27;Yan, 2012 #37;Qi, 2015 #47}two studies from Taiwan, China [14,15], one studies from Korea [17] and one study from worldwide [18]. The included studies were published between 2010 and 2020.The sample size for each study ranged from 41 to 503.The mean age of the patients was 46.96 years (ranged from 38.3 to 55.2years old). The duration of follow-up ranged from 1 year to 5 year.Male patients accounted for approximately 70.2% (ranged from 66.7% to 77%) of all patients.Two publications [14,15] reported by the same rst author were con rmed not to overlap due to the use of different study periods,so they were included in the study.

Changes in eGFR with LdT therapy
Six included studies,involving 514 patients, reported the change in eGFR after 1 year of LdT treatment (Fig. 2). The results showed that the eGFR was improved by 7.02 mL/min/1.73 m2 with LdT after 1 year of treatment.There was statistical heterogeneity was observed among these studies(I 2 = 85%,P < 0.00001) with a random-effects model.The result indicated a statistically signi cant change in the eGFR signi cant (RR = 7.02, 95%CI: 2.69-11.35, Z = 3.17; P =0.001).

Changes in eGFR with ETV therapy
Five included studies, including 546 patients,investigated the change in eGFR after 12 months of treatment with ETV (Fig. 3).The results showed that the eGFR was slightly increased after ETV treatment compared with baseline (1.72 mL/min/1.73 m2), and there was no signi cant heterogeneity among these studies (I 2 = 43%,P = 0.14) with a xed-effect model.The overall test result indicated that the eGFR was slightly increased after 12 months of treatment with ETV (RR = 1.72, 95%CI: 0.09-3.35,Z = 2.07; P = 0.04).

Renal safety comparison between LdT therapy group and ETV therapy group
There is no signi cant differences in baseline demographic data (age, gender) and eGFR level between the study groups.Two studies comprising 412 patients reported the change in eGFR after 6 months of treatment,as shown in Figure 4a,The eGFR was signi cantly higher in the LdT therapy group than in the ETV group after 6 months(RR = 4.63, 95%CI: 0.73-8.54,Z = 2.33; P = 0.02).and there was no signi cant heterogeneity among these studies (I 2 = 0%,P = 0.95) with a xed-effect model. Five studies comprising 788 patients reported the change in eGFR after 1 year of treatment,as shown in Figure 4b,The eGFR was signi cantly higher in the LdT therapy group than in the ETV group after 1 year(RR = 3.35, 95%CI: 1.18-5.52,Z = 3.02; P = 0.002).and there was no signi cant heterogeneity among these studies (I 2 = 0%,P = 0.82) with a xed-effect model.Three studies comprising 284 patients reported the change in eGFR after 2 year of treatment,as shown in Figure 4c,The eGFR was signi cantly higher in the LdT therapy group than in the ETV group after 2 year(RR = 11.00, 95%CI: 4.84-17.15,Z = 3.50; P = 0.0005) .and there was signi cant heterogeneity among these studies (I 2 = 53%,P = 0.12) with a random-effects model. The changes trends in eGFR during treatment in the LdT and ETV groups are shown in Fig.5.

Risk of bias
All trials were evaluated by the Cochrane Collaboration's risk-of-bias tool.The risk of bias assessment conducted for each study included is presented in Fig. 6.

Discussion
Renal function is an important prognostic factor in CHB patients.Amet et al reported that 64.6% of patients were found to have renal abnormalities by international de nition [20].The study by Raquel et al demonstrated that the percentage of Renal and urinary disorders with LAM,ETV,LdT,ADV,TDF,and TAF were 0.02%1.6%,0.1%,0.4%,6.8%and 11.1% [21].The risk factors for renal abnormalities include aging,gender,smoking,alcohol intake,diabetes,hypertension,anaemia and dyslipidaemia [20,22,23].so they recommend appropriate on-treatment monitoring of renal function [20].All currently available NAs are primarily predominantly eliminated unchanged in urine.These oral Antiviral agents,especially the nucleotide analogs,are associated with a dose-dependent nephrotoxicity,and both of pre-existing renal insu ciency and concomitant nephrotoxic agents are considered to be the risk factors of nephrotoxicity [24].Therefore,Special attention should be given to patients with pre-existing renal insu ciency who have been treated with Nucleos(t)ide Analogues,because they have a high tendency to develop renal dysfunction during prolonged CHB therapy [25].
We found that there's signi cant difference in comparing telbivudine with entecavir from 6 months to two years follow up period.Previous studies indicated that the improvement in eGFR was maintained longterm in telbivudine therapy [16,18].However,the mechanism by which telbivudine therapy improves renal function is still under investigation,it perhaps due to suppress ACE levels,which can control reninangiotensin aldosterone regulatory system and affect systemic vasoconstriction and renal sodium and renal uid retentions. Liang et al. reported that after about 1 year telbivudine treatment,Patients' eGFR was found signi cant increase.The serum angiotensin converting enzyme levels were negatively correlated with eGFR (r = − 0.375, p = 0.002).Signi cant decreases of the serum angiotensin converting enzyme levels were also observed upon entecavir treatment,but no signi cant correlation was found between serum serum angiotensin converting enzyme levels and eGFRs (r = -0.239, P = 0.138) [26].
Some previous studies reporting that there was no obvious difference in mean eGFR among patients treated with entecavir [24,27],some studies indicated that eGFR in CHB patients improved signi cantly after entecavir treatment,specially in renal patients [14,15].Mandíková J demonstrates that the potency of ETV to cause nephrotoxicity and/or clinically signi cant drug-drug interactions related to the tested transporters is considerably lower than that of nucleotide analogs [9].One thing for sure is that the ETVtreated did not deteriorate signi cantly compared to baseline.
Tenofovir disoproxil fumarate (TDF),a prodrug of tenofovir,has been shown to have a potential nephrotoxic [25] There is also study found that patients treated with TDF were not associated with renal impairment than patients treated with entecavir,but pre-existing renal insu ciency can increase the risk of developing changes in renal function [28].However,tenofovir alafenamide(TAF),a new tenofovir salt formulation,was shown to have better renal and bone safety than TDF [29],can be used as a replacement drug.Patients who have been treated with TDF,especially in who with a baseline eGFR of below 90 mL/min,can swich TAF to improve renal function [29,30].However,the number of patients receiving with TAF is too small to consolidate that TAF has a less impact on renal function than TDF.
There were none of network Meta-analysis published focusing on renal safety of antiviral therapy of CHB patients with renal impairment.Also,there were several limitations in this study.First,the numbers of studies were modest,only seven researches were included in the meta-analysis,and the number of prospective studies is also limited.Second,most studies were retrospective cohort studies and observational,with only two RCT, so there is an increased selection bias risk in retrospective studies.Third,eGFR was evaluated after only 2 year. As the duration of treatment of patients with CHB is several years, the clinical signi cance of the reported eGFR changes needs further elucidation.Finally, the participants covered in our meta-analysis are mainly from Asian countries,where the prevalence of HBV infection is high,and this might limit the generalizability of the results to multiple ethnicities.

Conclusions
In conclusion,this meta-analysis current evidence demonstrated that compared with ETV therapy,LdT has a signi cant improvement in eGFR in CHB patients with renal impairment.Patients with renal impairment in particular bene ted from telbivudine therapy.So in these patiens,LdT could be the better choice than ETV.    The change in eGFR after 1 year of entecavir treatment