This review concluded that adefovir dipivoxil and pegylated interferon alpha-2a are associated with significant improvements in a number of biochemical, virological and histological outcomes in adults with chronic hepatitis B infection. Whilst this was a well-conducted systematic review, the generalisability of the review's findings to patient subgroups routinely encountered in clinical practice is unclear.

To assess the effectiveness of adefovir dipivoxil (ADV) and pegylated interferon alpha-2a (PEG-IFN) for the treatment of adults with chronic hepatitis B (CHB) infection.

The Cochrane Database of Systematic Reviews, the Cochrane CENTRAL Register, DARE, HTA, NHS EED, MEDLINE, PREMEDLINE, EMBASE, EconLit, the National Research Register, the Science Citation Index (via ISI Web of Science), ISI Proceedings, BIOSIS Previews, ClinicalTrials.gov and Current Controlled Trials were searched from 1995/6 to April 2005; the search terms were reported. The searches were limited to the English language. The authors also searched several named relevant websites. To identify additional relevant studies, the authors screened the reference lists of related papers and manufacturer and sponsor submissions to the National Institute for Health and Clinical Excellence, and contacted experts.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) were eligible for inclusion. Most of the included studies were multicentre RCTs, conducted in several countries across Europe, Asia, North America and Australasia. The included studies were mostly funded by drug manufacturers.

Specific interventions included in the review

Studies that compared ADV or PEG-IFN (alone or in combination with other treatments) with currently licensed treatments for CHB, including non-pegylated interferon alpha (IFN), lamivudine (LAM) and best supportive care (BSC), or placebo, were eligible for inclusion. The included studies of ADV used dosages of 10 to 30 mg/day (most used 10 mg/day), studies of PEG-IFN used dosages of 90 to 270 microg/week (most used 180 microg/week), studies of LAM used dosages of 100 mg/day, and studies of IFN used dosages of 4.5 MIU three times per week.

Participants included in the review

Studies of adults with CHB were eligible for inclusion. The majority of included studies were of patients with hepatitis B e antigen (HBeAg)-positive CHB; the remaining trials were of patients with HBeAg-negative CHB. None of the studies included patients co-infected with the human immunodeficiency virus, and most studies also excluded patients co-infected with hepatitis C virus or hepatitis D virus. The average age of the participants ranged from 31 to 46 years and 74 to 95% were male.

Outcomes assessed in the review

The short-term outcomes of interest were biochemical, histological and virological response to treatment, seroconversion, drug resistance and adverse effects. The long-term outcomes of interest were survival, time to treatment failure and health-related quality of life (HRQoL). The majority of the included studies reported outcomes at the end of 48 weeks' treatment; none of the included studies reported longer term outcomes of survival or time to treatment failure. HRQoL was measured using the Short Form 36 questionnaire.

How were decisions on the relevance of primary studies made?

One reviewer screened the titles and abstracts of identified studies, whilst a second reviewer checked a random sample of 10%. Two independent reviewers assessed the full text versions of potentially relevant papers. Any disagreements were resolved by discussion.

Validity was assessed on the basis of the following: randomisation; allocation concealment; similarity of the groups at baseline; reporting of eligibility criteria; blinding; whether point estimates and measure of variability were presented for the primary outcome measure; whether an intention-to-treat-analysis was undertaken; and whether withdrawals and drop-outs were completely described. The validity of the unpublished study was not assessed.

One reviewer performed the validity assessment and a second reviewer checked it. Any disagreements were resolved by discussion.

One reviewer performed the data extraction using a standardised template, and a second reviewer checked it. Any disagreements were resolved by discussion.

How were the studies combined?

The studies were combined in a narrative.

How were differences between studies investigated?

Study details were presented in tables and text. The authors stated that the studies were too heterogeneous for statistical pooling.

Eight RCTs (n=2,529) were included in the review, although one was only reported as a conference abstract.

None of the included published RCTs reported the method of randomisation and only two reported adequate concealment of allocation. Baseline participant characteristics were reported, with none of the RCTs reporting any significant differences between study groups. Blinding of the outcome assessors was adequate in the majority of RCTs. The primary outcomes were adequately reported in all of the included published RCTs, but only 2 RCTs reported an adequate intention-to-treat analysis. Withdrawals were only fully described in 3 RCTs.

Clinical effectiveness of ADV (5 RCTs, including one unpublished).

The 4 published RCTs reported statistically significantly greater virological response and biochemical response in the ADV group in comparisons between ADV versus placebo, ADV plus LAM versus LAM alone, and ADV alone and ADV plus LAM versus LAM alone. Two RCTs comparing ADV with placebo reported statistically significantly greater histological improvement in the ADV groups than in the placebo groups.

ADV was significantly more effective than placebo in treatment-naive patients. There were no statistically significant differences in HBeAg loss or seroconversion rates between ADV, ADV plus LAM, or ongoing LAM in patients with resistance to LAM. HBeAg loss or seroconversion was observed in less than 5% of patients taking ADV.

Drug resistance rates were generally 5-fold lower with ADV than with LAM. After 4 years of treatment the cumulative rates were 14.5% and 70%, respectively.

Dose discontinuations for safety reasons were low for patients receiving ADV. Headache was the only adverse event consistently seen more frequently in the ADV groups than in the placebo groups.

Clinical effectiveness of PEG-IFN (3 RCTs).

Virological and biochemical response rates were similar for patients treated with PEG-IFN monotherapy and PEG-IFN plus LAM. Both groups had statistically significantly higher virological and biochemical response rates than patients treated with LAM monotherapy. There was no statistically significant difference in virological or biochemical response rates between PEG-IFN and IFN. Two RCTs comparing PEG-IFN monotherapy versus PEG-IFN plus LAM versus LAM monotherapy found no statistically significant differences in histological response between any of the groups.

HBeAg loss or seroconversion rates were significantly higher for patients treated with PEG-IFN monotherapy and PEG-IFN plus LAM than for those receiving LAM monotherapy. Differences between PEG-IFN and IFN were not statistically significant.

HRQoL was reported in 2 studies of PEG-IFN; the scores decreased during treatment, but returned to at least baseline levels at follow-up (based on unpublished data). For HBeAg-positive patients there were no significant differences in scores between groups receiving PEG-IFN and those receiving LAM. For HBeAg-negative patients, scores were significantly better in the PEG-IFN plus LAM group compared with the LAM monotherapy group for two of the HRQoL components.

Dose discontinuations for safety reasons were significantly higher for patients receiving PEG-IFN than for those receiving LAM monotherapy. The most commonly reported adverse events in the PEG-IFN studies were headache, pyrexia, fatigue, myalgia and alopecia; these were all more commonly seen in patients receiving PEG-IFN than in those receiving LAM monotherapy.

A systematic review of cost-effectiveness studies was undertaken and an economic model was developed to estimate the cost-effectiveness of PEG-IFN and ADV compared with IFN, LAM and BSC for adults with CHB. The incremental cost per quality-adjusted life-year estimates were £5,994 for IFN compared with BSC, £6,119 for PEG-IFN compared with IFN, £3,685 for LAM compared with BSC and £16,569 for ADV compared with LAM.

ADV and PEG are associated with significant improvements in a number of biochemical, virological and histological outcomes in both HBeAg positive and HBeAg negative CHB patients; this is associated with resolution of infection for a small proportion of patients. For another proportion of patients it leads to remission and a reduced risk of progression to cirrhosis, hepatocellular carcinoma, liver transplant and death. For others who do not respond or who relapse, re-treatment with another agent is necessary. The severity and frequency of serious adverse events associated with treatment appears relatively low.

The review question was clear in terms of the study designs, participants, interventions and outcomes of interest. Several relevant electronic databases were searched and the search terms were reported. The authors sought unpublished data, thereby reducing the potential for publication bias. However, only studies reported in English were eligible for inclusion, thus increasing the potential for language bias. Reasonable attempts were made to reduce error and bias in the screening studies for relevance, validity assessment and data extraction processes. The criteria used for the quality assessment were appropriate. Some quality criteria were either not reported or not met by the included studies, which may have implications for the validity of the conclusions of the review. Adequate study details were presented; the narrative synthesis was appropriate in view of the differences between the studies.

Overall, the conclusions of this well-conducted systematic review appear to follow from the evidence presented. The generalisability of the findings of the review to patient subgroups routinely encountered in clinical practice is unclear; the authors stated that further research is needed to assess the effectiveness of antiviral treatment for CHB infection in different subgroups.

Practice: The authors stated that there appears to be no significant barriers to diffusion of the appraised treatments into routine practice; both PEG and ADV are in current use, to varying extents. However, policy-makers need to view the evidence for clinical and cost-effectiveness within the wider context of hepatitis B, taking into consideration primary prevention, vaccination, screening and investigation, and the changing epidemiology of infection in England and Wales.

Research: The authors stated that further research, in the form of RCTs, should be conducted to assess the effectiveness of antiviral treatment for CHB infection, particularly for subgroups of patients with different genotypes, patients with cirrhosis, patients from different ethnic groups, patients with co-infections (e.g. human immunodeficiency virus, hepatitis C virus) and co-morbidities, liver transplant patients, and children and adolescents.

NHS R&D Health Technology Assessment (HTA) Programme, project number 03/59/01.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.