Review
Treatment outcomes after highly active antiretroviral therapy: a meta-analysis of randomised controlled trials

https://doi.org/10.1016/S1473-3099(04)01057-6Get rights and content

Summary

This systematic review summarises the evidence for treatment efficacy and tolerability of highly active antiretroviral therapies containing two nucleoside reverse transcriptase inhibitors (NRTI) with a protease inhibitor (PI), compared with two NRTIs alone for the treatment of HIV-1 infection in randomised controlled trials. Three electronic databases (Medline, Embase, and the Cochrane Library) were searched up to December 2003. 16 randomised controlled trials met the inclusion criteria and were included in the analysis from 328 articles screened. The pooled analysis indicated that treatment with two NRTIs with a PI is more effective in achieving viral suppression than two NRTIs alone (relative risk [RR] 3·44, 95% confidence interval [CI] 2·43–4·87). However, the RR for discontinuation of treatment due to adverse events of treatment with two NRTIs with a PI compared with two NRTIs alone was 1·81 (95% CI 1·17–2·79). The benefits of treatment with two NRTIs and a PI are substantial among those who can tolerate the regimen in comparison with treatment with two NRTIs alone.

Section snippets

Search strategy and selection criteria

The following electronic databases were searched for randomised controlled trials of HIV-1 infection published up to December 31,2003: PubMed (Medline, 1966–2003), Embase (1993–2003), and the Cochrane Controlled Trials Register Database (Cochrane Library, Issue 1, 2004). As of June 2003, there were seven protease inhibitors that were licensed for the treatment of HIV-1-infected adults and adolescents in the USA.9 Separate search strings were developed for HIV/AIDS, HAART, and randomised

Primary outcomes

The primary outcomes of this review were the risk of discontinuation of assigned treatment due to adverse events, discontinuation of assigned treatment due to any cause, death, treatment efficacy as defined by each of the studies, and mean decrease in HIV-1 RNA. Since reductions in plasma viraemia achieved with antiretroviral therapy account for substantial clinical benefits,16 treatment efficacy was defined in terms of suppression of viral load or mean decrease in plasma HIV-1 RNA. The risk

Data abstraction

From each trial report, two independent reviewers abstracted characteristics of the trial: year of publication, the location of the study, whether or not it was a multicentre trial, previous treatment with NRTIs, NNRTIs, or PIs, length of follow-up, and the individual drugs comprising the treatment groups (drug, dose, and frequency). The followup time for trials in which HIV-infected adults who had successfully completed antiretroviral induction therapy were randomised to maintenance therapy

Statistical analysis

Analyses were done using Stata 8·0. Overall effect measures were calculated using pooled risk ratios (RR) with 95% confidence intervals (CIs) for the outcomes of interest according to the Mantel-Haenszel fixed effects22 and DerSimonian-Laird random effects23 models. The weighted average of differences of mean change of HIV-1 RNA (log copies/mL) was calculated using an unstandardised mean difference. Between-study heterogeneity was assessed using the χ2 statistic.24 The random effects model is

Results

The search strategy yielded 118 studies that satisfied the inclusion/exclusion criteria from the 328 full-text articles evaluated (figure 1). From these 118 studies, the number of treatment comparisons examined in each report were tallied and grouped into their respective class of treatments. This review focused on the most frequently studied comparison—ie, two NRTIs + PI versus two NRTIs (25 comparisons from 19 studies). Three studies26, 27, 28 used the same data for two other primary studies29

Treatment efficacy

Treatment efficacy was defined by each of the trials as HIV RNA below some cut-off value (copies/mL) by the end of study (table 3 and figure 2). The pooled analysis showed that treatment consisting of two NRTIs plus PI is more effective in achieving viral suppression than two NRTIs without a PI (RR=3·44, 95% CI 2·43, 4·87). In the subgroups analysis (table 4), treatment with zidovudine, lamivudine, and nelfinavir increased the risk for viral suppression in comparison to treatment with

Mean change in HIV-1 RNA

The greatest weighted mean difference in mean change of HIV-1 RNA (log copies per mL) was between treatment with zidovudine, lamivudine, and indinavir versus zidovudine and lamivudine (difference in mean change= −1·15, 95% CI −1·17, −1·13, table 5). A pooled weighted mean difference across the seven comparisons that reported mean change in HIV-1 RNA was −0·61 (95% CI −0·94, −0·29), further supporting the efficacy of including a PI with two NRTIs in decreasing plasma viral load.

Discontinuation of treatment due to adverse events

17 of 22 possible comparisons reported information on withdrawal due to adverse events (table 6 and figure 3). People treated with two NRTIs and a PI were more likely to withdraw due to adverse events than those who were treated with two NRTIs (the DerSimonian-Laird weighted pooled risk ratio assuming random effects was 1·81, 95% CI 1·17–2·79). There was strong evidence to reject the null hypothesis that the effect estimates were homogeneous (test for heterogeneity χ2=43·36, degrees of freedom

Discontinuation of treatment due to all causes

By contrast, there were 11 comparisons of 22 possible that reported information on total withdrawals from assigned treatment due to all causes (table 7 and figure 4). The DerSimonian-Laird weighted pooled risk ratio assuming random effects was 0·70, 95% CI 0·56–0·88 suggesting that those who are taking two NRTIs plus PI are at a decreased risk of withdrawing from assigned treatment due to all causes. The test for heterogeneity determined that the effects were heterogeneous (χ2=21·41, df=10,

All-cause mortality

The 15 comparisons that reported information on deaths by treatment group are given in table 8. Nine of these comparisons did not have any deaths for either of the treatment groups or the comparison groups and were thus excluded from the analysis. Among the six comparisons that had deaths in either treatment group, there was evidence that treatment with two NRTIs plus PI decreased the risk for death in comparison with treatment with two NRTIs without a PI (RR=0·64, 95% CI 0·46, 0·88). A

Adverse events reported among the trials

Among the clinical adverse events reported for each treatment group, nausea and diarrhoea were the most frequently reported (five studies) followed by headache and rash (four studies). Among the laboratory adverse events reported for each treatment group, three studies reported the number of people that developed anaemia by treatment group. Among the nine studies (12 comparisons) that reported new AIDS-defining events by treatment group, four studies (five comparisons) did not have any events (

Investigation of publication bias

There was evidence for bias according to the Egger's weighted regression approach (p=0·001)46 as would be expected in a review that only includes published trials since statistically significant (“positive”) studies are more likely to be submitted and accepted for publication.47

Discussion

This systematic review of PI-containing HAART therapy for treatment of HIV-1 infection showed a consistent and significantly greater benefit of triple therapy as measured by suppressing viral load in comparison with double therapy that does not contain a PI. When the effects of individual comparisons of treatments reported among the included trials were investigated, there was considerable heterogeneity in the effect of including a protease inhibitor, depending on the base treatment of NRTIs

Search strategy and selection criteria

These are described in detail in the Methods section on page 412.

References (65)

  • A Carr

    Improvement of the study, analysis, and reporting of adverse events associated with antiretroviral therapy

    Lancet

    (2002)
  • JP Ioannidis et al.

    The impact of high-risk patients on the results of clinical trials

    J Clin Epidemiol

    (1997)
  • SL Aversa et al.

    Psychosocial aspects of antiretroviral medication use among HIV patients

    Patient Educ Couns

    (1996)
  • RS Hogg et al.

    Improved survival among HIV-infected individuals following initiation of antiretroviral therapy

    JAMA

    (1998)
  • Jr Palella FJ et al.

    Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators

    N Engl J Med

    (1998)
  • M Dybul et al.

    Guidelines for using antiretroviral agents among HIV-infected adults and adolescents. Recommendations of the Panel on Clinical Practices for Treatment of HIV

    MMWR Recomm Rep

    (2002)
  • HK Thaker et al.

    HIV viral suppression in the era of antiretroviral therapy

    Postgrad Med J

    (2003)
  • GR Kaufmann et al.

    Immune reconstitution in HIV-1 infected subjects treated with potent antiretroviral therapy

    Sex Transm Infect

    (1999)

  • Report of the NIH Panel to Define Principles of Therapy of HIV Infection

    MMWR Recomm Rep

    (1998)

  • 1999 USPHS/IDSA guidelines for the prevention of opportunistic infections in persons infected with human immunodeficiency virus

    Clin Infect Dis

    (2000)

  • Panel on Clinical Practices for Treatment of HIV Infection

    Guidelines for the use of antiretroviral agents in HIV-1 infected adults and adolescents

    (2003)
  • R Detels et al.

    Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration. Multicenter AIDS Cohort Study Investigators

    JAMA

    (1998)
  • K Porter et al.

    Determinants of survival following HIV-1 seroconversion after the introduction of HAART

    Lancet

    (2003)
  • KA Robinson et al.

    Development of a highly sensitive search strategy for the retrieval of reports of controlled trials using PubMed

    Int J Epidemiol

    (2002)
  • S Derry et al.

    Incomplete evidence: the inadequacy of databases in tracing published adverse drug reactions in clinical trials

    BMC Med Res Methodol

    (2001)
  • JP Ioannidis et al.

    Meta-analysis of randomised controlled trials on the comparative efficacy and safety of azithromycin against other antibiotics for upper respiratory tract infections

    J Antimicrob Chemother

    (2001)
  • WA O'Brien et al.

    Changes in plasma HIV-1 RNA and CD4+ lymphocyte counts and the risk of progression to AIDS

    N Engl J Med

    (1996)
  • R Jordan et al.

    Systematic review and meta-analysis of evidence for increasing numbers of drugs in antiretroviral combination therapy

    BMJ

    (2002)
  • KF Schulz et al.

    Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials

    JAMA

    (1995)
  • JP Ioannidis et al.

    Completeness of safety reporting in randomised trials: an evaluation of seven medical areas

    JAMA

    (2001)
  • N Mantel et al.

    Statistical aspects of the analysis of data from retrospective studies of disease

    J Natl Cancer Inst

    (1959)
  • J Lau et al.

    Quantitative synthesis in systematic reviews

    Ann Intern Med

    (1997)

    • Cited by (20)

    • Dyslipidemia in AIDS patients on highly active antiretroviral therapy

      2011, Brazilian Journal of Infectious Diseases

    • Meta-analytical Studies on the Epidemiology, Prevention, and Treatment of Human Immunodeficiency Virus Infection

      2009, Infectious Disease Clinics of North America
      Citation Excerpt :

      Results of the immunologic and virologic response were consistent with the clinical outcomes. Another meta-analysis of multicenter RCTs addressed the outcome of treatment using two different regimens.26 The pooled analysis showed that treatment containing two non-nucleoside reverse transcriptase inhibitors (NNRTIs) and a protease inhibitor (PI) was consistently more effective in suppressing HIV-1 viral load compared with two NRTIs alone (RR 3.44, 95% CI, 2.43–4.87).

    • Initial experience with bariatric surgery in asymptomatic human immunodeficiency virus-infected patients

      2005, Surgery for Obesity and Related Diseases

    • Evidence-based clinical guidelines for immigrants and refugees

      2011, CMAJ. Canadian Medical Association Journal
      Citation Excerpt :

      We were unable to find quantitative data on anxiety or depression, changes in relationships with sexual partners or discrimination associated with false-positive results. A meta-analysis by Enanoria and colleagues256 comparing three-drug and two-drug antiretroviral treatment showed a significant decrease in mortality (risk ratio 0.62, 95% confidence interval 0.45–0.86) (Table 8C). In another meta-analysis, Siegfried and coauthors247 reported a relative risk of 1.3 for adverse events associated with antiretroviral treatment.

    • Evaluation of evidence-based literature and formulation of recommendations for the clinical preventive guidelines for immigrants and refugees in Canada

      2011, CMAJ. Canadian Medical Association Journal
      Citation Excerpt :

      Review group teams synthesized the evidence from the updated systematic reviews explicitly incorporating immigrant and refugee population-specific clinical considerations and value judgments in order to draft preferably no more than two key clinical actions, targeting (where necessary) specific populations or regions. Both desirable and undesirable effects of the intervention were summarized in both absolute and relative terms for each patient-important outcome using the summary of findings table format adopted by the Cochrane Collaboration12 (Table 2).13 The quality for each outcome was assessed using the items specified by the GRADE Working Group (indirectness, consistency, precision, reporting bias and study limitations).

    • Dyslipidemia and associated risk factors among HIV/AIDS patients on HAART in Asmara, Eritrea

      2022, PLoS ONE
    View all citing articles on Scopus
    View full text
    Copyright © 2004 Elsevier Ltd. All rights reserved.