|A safety review and meta-analyses of bevacizumab and ranibizumab: off-label versus goldstandard
|Schmucker C, Ehlken C, Agostini HT, Antes G, Ruecker G, Lelgemann M, Loke YK
This review concluded that it was not possible to rule out a clinically relevant risk of serious adverse effects with unlicensed bevacizumab. Due to limited data and concerns about study quality, the conclusions seem appropriately cautious and are likely to be reliable.
To evaluate whether unlicensed therapy with bevacizumab was as safe as licensed therapy with ranibizumab, and whether clinicians were justified in offering bevacizumab to their patients with age-related macular degeneration, with no additional risk.
MEDLINE, EMBASE and The Cochrane Library were searched up to May 2011 with no language or date limitations. Search terms were reported. Trials registries and references of included studies and previous reviews were consulted. The manufacturer of both drugs (Genentech) was contacted for additional data.
Randomised controlled trials (RCTs) that investigated bevacizumab or ranibizumab, in direct comparisons or against any other control group, in patients with neovascular age-related macular degeneration, were eligible. Trials that compared different regimens of ranibizumab or bevacizumab were included. Follow-up had to be at least one year. Trials of less than 20 patients were excluded.
In the included trials, ranibizumab and bevacizumab were compared against each other or against various active or inactive treatments. Treatment was generally administered monthly or as needed. The ranibizumab dosage was either 0.3mg or 0.5mg and the bevacizumab dosage was either 1mg or 1.25mg. The mean number of injections per patients varied widely.
Study selection was performed independently by two reviewers, with disagreements resolved by discussion and consensus.
Assessment of study quality
Data extraction and quality assessment were carried out using a modified evaluation tool from the Centre for Reviews and Dissemination. Quality criteria covered: comparability of groups, blinding, definition of expected adverse event, definition of method to collect adverse event data, and transparency of patient flow, with an overall assessment of low, moderate, or high.
Quality assessment was performed independently by two reviewers, with disagreements resolved by discussion and consensus.
Safety outcomes (ocular and non-ocular adverse events) were extracted to calculate risk ratios and 95% confidence intervals. Data extraction was performed independently by two reviewers, with disagreements resolved by discussion and consensus.
Methods of synthesis
Where possible, a meta-analysis using a fixed-effect model (Mantel-Haenszel method) was employed to pool the safety outcomes, with heterogeneity assessed using Χ² and Ι². Otherwise, a narrative summary was reported.
Results of the review
Eleven RCTs, with 5,631 patients, were included. Three were head-to-head trials comparing ranibizumab with bevacizumab (28 to 1,185 patients), with follow-up ranging from 12 to 18 months. Several gaps in the quality of the head-to-head trials were reported. None reported complete comparability between groups. Adequate blinding, the definition of an expected adverse event, or a method to collect adverse events were adequately reported in only one trial. Patient dropout was unclear in all three RCTs.
The risk of ocular adverse events was significantly higher for patients receiving bevacizumab, compared with ranibizumab (RR 2.8, 95% CI 1.2 to 6.5; three RCTs). The risk of serious systemic adverse effects was significantly higher with bevacizumab than with ranibizumab (RR 1.3, 95% CI 1.0 to 1.7; one RCT). Rates of death (two RCTs) and serious non-ocular haemorrhage (one RCT) were higher in patients receiving ranibizumab, but the differences, compared with bevacizumab, were not statistically significant. There was no significant difference in the rates of arterial thromboembolic events.
Trials comparing bevacizumab with control reported few adverse events in the intervention group and no significant difference compared with control (two RCTs; 238 patients). Safety results comparing ranibizumab with control treatment were reported.
Evidence from head-to-head trials indicated an increased risk of ocular and multiple systemic adverse events with bevacizumab. It was unclear whether this was due to genuine differences in systemic toxicity, or differences in the drug's manufacture.
The review question and inclusion criteria were clearly stated and reproducible. Several sources were searched, with no language or date restrictions. Appropriate steps to minimise reviewer error and bias were taken throughout the review. A relatively large number of trials was included in the review, but only three directly compared the two treatments. All had methodological limitations. Only one was large, and so this significantly influenced the direction of the pooled analyses. The method of synthesis appears to have been appropriate, given the differences between the trials.
Due to limited data on the safety of bevacizumab, the conclusions seem appropriately cautious and are likely to be reliable.
Implications of the review for practice and research
Practice: The authors stated that clinicians and patients should carefully assess the benefits and harms when choosing between the two options. They noted a need for high surveillance for systemic adverse effects, with intraocular anti-vascular endothelial growth factor injections, for age-related macular degeneration and other retinal diseases.
Research: The authors stated that the results from ongoing trials should help to clarify whether the increased risks of adverse effects were related to intravitreal anti-vascular endothelial growth factor therapy. They stated that studies should be powered not just for efficacy, but for defined safety outcomes, based on the results of this review.
Supported by a grant from the German Federal Ministry of Education and Research.
Schmucker C, Ehlken C, Agostini HT, Antes G, Ruecker G, Lelgemann M, Loke YK. A safety review and meta-analyses of bevacizumab and ranibizumab: off-label versus goldstandard. PLOS ONE 2012; 7(8): e42701
Subject indexing assigned by NLM
Antibodies, Monoclonal, Humanized /adverse effects /pharmacology; Bias (Epidemiology); Dose-Response Relationship, Drug; Eye /drug effects /pathology; Humans; Off-Label Use /standards; Randomized Controlled Trials as Topic; Treatment Outcome
Date bibliographic record published
Date abstract record published
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.