A systematic review and meta-analysis of erythropoietin in experimental stroke

Abstract

Erythropoietin (EPO) has shown promise as a neuroprotectant in animal models of ischemic stroke. EPO is thought not only to protect neurons from cell death, but also to promote regeneration after stroke. Here, we report a systematic review and meta-analysis of the efficacy of EPO in animal models of focal cerebral ischemia. Primary outcomes were infarct size and neurobehavioral outcome. Nineteen studies involving 346 animals for infarct size and 425 animals for neurobehavioral outcome met our inclusion criteria. Erythropoietin improved infarct size by 30.0% (95% CI: 21.3 to 38.8) and neurobehavioral outcome by 39.8% (33.7 to 45.9). Studies that randomized to treatment group or that blinded assessment of outcome showed lower efficacy. Erythropoietin was tested in animals with hypertension in no studies reporting infarct size and in 7.5% of the animals reporting neurobehavioral outcome. These findings show efficacy for EPO in experimental stroke, but when the impact of common sources of bias are considered, this efficacy falls, suggesting we may be overestimating its potential benefit. As common human co-morbidities may reduce therapeutic efficacy, broader testing to delineate the range of circumstances in which EPO works best would be beneficial.

Figure 1

Effect on the estimate of efficacy for infarct volume of (A) type of delivery, (B) method of artery occlusion, and for neurobehavioral outcome of (C) type of delivery, (D) method of artery occlusion, and (E) model of ischemia, P<0.003. The shaded gray bar represents the 95% confidence limits of the global estimate. The vertical error bars represent the 95% confidence intervals for the individual estimates. The width of each bar reflects the log of the number of animals contributing to that comparison.

Figure 2

Effect on the estimate of efficacy for infarct volume of (A) random allocation to treatment or control group, and for neurobehavioral outcome of (B) blinded assessment of outcome, P<0.003.

Figure 3

Effect on the estimate of efficacy of quality score (1=worst, 11=best) for (A) infarct volume and (B) neurobehavioral outcome, P<0.003.