|Standardized treatment of active tuberculosis in patients with previous treatment and/or with mono-resistance to isoniazid: a systematic review and meta-analysis
|Menzies D, Benedetti A, Paydar A, Royce S, Pai M, Burman W, Vernon A, Lienhardt C
This review assessed outcomes related to treatment failure in tuberculosis patients who had been previously treated or who had documented isoniazid mono-resistance. The authors concluded that there was little published evidence to support the continued use of the currently recommended regimen for the treatment of these patient groups. This conclusion is probably reliable.
To assess treatment of tuberculosis in patients with a history of previous treatment, or with documented isoniazid mono-resistance. Factors associated with these outcomes were also investigated.
PubMed, Cochrane Central Register of Controlled Trials (CENTRAL) and EMBASE were searched from 1965 to June 2008 for studies published in English, French or Spanish. Search terms were reported. References of identified studies, reviews, chapters of textbooks and recent treatment guidelines were also checked.
Randomised controlled trials (RCTs) in patients with bacteriologically confirmed active pulmonary tuberculosis, that reported bacteriologically confirmed treatment failure or relapse, were eligible for inclusion. Treatment was required to be standardised and to include rifampin. Studies that used rifapentine, rifabutin or non drug therapies were excluded, as were regimens with once weekly or single drug therapies. Cohort studies that evaluated the World Health Organisation's (WHO) re-treatment regimen were also eligible for inclusion, provided that they reported individual outcomes and met other inclusion criteria. Studies identified in a previous systematic review (see Other Publications of Related Interest), in which treatment-naive patients with isoniazid mono-resistance were treated with standardised rifampin-based regimens, were also included.
The outcomes were failure, relapse and acquired drug resistance; definitions of outcomes were reported in the paper.
Regimens used in included studies involved combinations of the following agents: isoniazid, rifampin, pyrazinamide, ethambutol and streptomycin. Most included studies used a 600mg dose of rifampin.
Two reviewers independently assessed the studies for inclusion in the review.
Assessment of study quality
Two reviewers independently assessed criteria including the number of withdrawals, drop-outs and refusals of re-treatment, and methods of randomisation relating to allocation concealment. Disagreements were resolved through consensus. It did not appear that cohort studies were assessed for validity.
Two reviewers independently extracted data on the numbers of patients who experienced each outcome. Disagreements were resolved through consensus.
Methods of synthesis
A short narrative synthesis of cohort studies of WHO standardised regimens and of RCTs in previously treated patients was provided. As there was very little direct head-to-head evidence, an indirect treatment comparison was performed using a binomial regression model, including study as a random-effect. A meta-regression was also carried out to explore factors which affected responses to treatment in RCTs.
Results of the review
The review included 33 RCTs of which 24 were in new cases. Of the 24 trials in new patients, randomisation was high quality in 16 and loss to treatment of less than 10% was reported in 18.
In all patients with isoniazid mono-resistance (33 trials including 24 in new cases), the pooled relapse rate was 10%. The pooled relapse rate for trials enrolling new cases was 15%.
Cohort studies (five studies) of the WHO standardised re-treatment regimen showed low failure rates (with a single exception). Where mono-resistance to isoniazid was reported, failure rates ranged from 18 to 44%. Where drug resistance patterns were not reported, the rates ranged from 9 to 45%.
RCTs in previously treated patients (nine RCTs) with isoniazid mono-resistance showed mixed results. Three studies using standard treatment for drug-sensitive tuberculosis had combined failure and relapse rates ranging from 29 to 70%. In four trials of rifampin plus ethambutol for at least 12 months, failure rates ranged from 4 to 23% and relapse rates from 0 to 27%. A single trial evaluated a regimen similar to current re-treatment practice with a failure rate of 1% and a relapse rate of 3%. Randomisation was considered to be high quality in eight of these trials; six had loss to treatment of less than 10%.
Meta-regression found that treatment outcomes were significantly worse if therapy was intermittent (two or three times weekly) or included fewer effective drugs in the initial intensive treatment phase. Relapse rates were lower with longer durations of rifampin or pyrazinamide. There was no association between relapse and the duration of post-treatment follow-up.
There was little published evidence to support the continued use of the currently recommended regimen for the treatment of previously treated tuberculosis patients or those with known mono-resistance to isoniazid
The review question and the inclusion criteria were clear, although it was apparent that the protocol was amended to allow for the inclusion of non-randomised studies. The authors searched three relevant databases and other sources. The inclusion of studies published in three European languages only may have increased the risk that relevant studies were omitted, or that language bias was introduced. The authors used methods designed to reduce reviewer bias and error at all stages of the review process.
The mix of narrative synthesis and meta-analytic methods appeared appropriate, although indirect comparisons are not as reliable as evidence from direct head-to-head comparisons.
The authors' conclusions accurately reflected the results of the view and are probably reliable.
Implications of the review for practice and research
Practice: The authors stated that a concerted international effort to substantially expand access to reliable drug sensitivity testing is required. They also stated that whilst awaiting the results of further trials (see recommendations for further research below) the standardised treatment regimen should be redesigned, at a minimum to adequately treat patients with isoniazid resistance.
Research: The authors stated that there is an urgent need for RCTs in tuberculosis patients with pre-treatment drug resistance in all forms, particularly those previously treated.
Canadian Institutes of Health Research; Fonds de la Recherche en Sante du Quebec; World Health Organisation (WHO).
Menzies D, Benedetti A, Paydar A, Royce S, Pai M, Burman W, Vernon A, Lienhardt C. Standardized treatment of active tuberculosis in patients with previous treatment and/or with mono-resistance to isoniazid: a systematic review and meta-analysis. PLoS Medicine 2009; 6(9):e1000150
Other publications of related interest
Menzies D, Bendetti A, Paydar A, Martin I, Royce S, Pai M, Vernon A, Lienhardt C, Burman W. Effect of duration and intermittency of rifampin on tuberculosis treatment outcomes: A systematic review and meta-analysis. PLoS Medicine 2009;6(9):e1000146.
Subject indexing assigned by NLM
Antitubercular Agents /therapeutic use; Cohort Studies; Drug Resistance, Bacterial; Drug Therapy, Combination; Humans; Isoniazid /therapeutic use; Randomized Controlled Trials as Topic; Recurrence; Rifampin /therapeutic use; Streptomycin /therapeutic use; Treatment Failure; Tuberculosis, Pulmonary /drug therapy
Database entry date
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.