Importance: Although intravitreal anti-vascular endothelial growth factor (VEGF) treatment represents the first-line therapy for many retinal diseases, the issue of their systemic safety is debatable.
Objectives: To assess whether intravitreal anti-VEGF therapy might be associated with increased risk of mortality and which variables are associated with the increase.
Data sources: PubMed, MEDLINE, and Embase databases, the Cochrane Library, and ClinicalTrials.gov were systematically searched from inception to May 6, 2019.
Study selection: Randomized clinical trials comparing intravitreal anti-VEGF treatment with control groups and with follow-up of at least 6 months were selected.
Data extraction and synthesis: Data were independently collected by 2 investigators. Meta-analyses were conducted using the frequentist and Bayesian methods. For the frequentist approach, random- and fixed-effects models were used, with random-effects models considered the primary technique. Odds ratios (ORs) with 95% CIs were computed. For the bayesian approach, uninformative and informative priors were used. Odds ratios with 95% credible intervals (CrIs) were computed. Meta-regression analyses were based on random-effects models.
Main outcomes and measures: The primary outcome measure was the all-cause death rate. Secondary outcomes included meta-regression analyses on the following variables: type of drug, number of injections, follow-up time, diagnosis, and cardiovascular risk.
Results: Of 2336 studies identified, 34 unique studies with 8887 unique participants were included in the present meta-analysis. For the frequentist analysis, fixed- and random-effects models yielded similar estimates (ORs, 1.34 [95% CI, 0.95-2.07; P = .09] and 1.34 [95% CI, 0.89-2.01; P = .17], respectively). For the Bayesian approach, noninformative and informative priors yielded similar results (ORs, 1.34 [95% CrI, 0.79-2.34; 0.13 probability of OR≤1.00] and 1.40 [95% CrI, 0.82-2.32; 0.11 probability of OR≤1.00], respectively). Meta-regression analyses showed the following risk for 1 injection more: frequentist OR of 1.12 (95% CI, 1.04-1.22; P = .005) and Bayesian OR of 1.06 (95% CrI, 0.98-1.15; 0.06 probability of OR≤1.00).
Conclusions and relevance: In this study, no difference was found in the mortality rate between intravitreal anti-VEGF treatment and control groups. Additional data seem warranted to determine whether the mortality rate is increased in patients receiving a greater number of injections.