Central Nervous System Electrical Stimulation for Neuroprotection in Acute Cerebral Ischemia: Meta-Analysis of Preclinical Studies

Abstract

Background and Purpose- Brain electrical stimulation, widely studied to facilitate recovery from stroke, has also been reported to confer direct neuroprotection in preclinical models of acute cerebral ischemia. Systematic review of controlled preclinical acute cerebral ischemia studies would aid in planning for initial human clinical trials. Methods- A systematic Medline search identified controlled, preclinical studies of central nervous system electrical stimulation in acute cerebral ischemia. Studies were categorized among 6 stimulation strategies. Three strategies applied different stimulation types to tissues within the ischemic zone (cathodal hemispheric stimulation [CHS], anodal hemispheric stimulation, and pulsed hemispheric stimulation), and 3 strategies applied deep brain stimulation to different neuronal targets remote from the ischemic zone (fastigial nucleus stimulation, subthalamic vasodilator area stimulation, and dorsal periaqueductal gray stimulation). Random-effects meta-analysis assessed electrical stimulation modification of final infarct volume. Study-level risk of bias and intervention-level readiness-for-translation were assessed using formal rating scales. Results- Systematic search identified 28 experiments in 21 studies, including a total of 350 animals, of electrical stimulation in preclinical acute cerebral ischemia. Overall, in animals undergoing electrical stimulation, final infarct volumes were reduced by 37% (95% CI, 34%-40%; P<0.001), compared with control. There was evidence of heterogeneity of efficacy among stimulation strategies (I²=93.1%, P_{heterogeneity}<0.001). Among the within-ischemic zone stimulation strategies, only CHS significantly reduced the infarct volume (27 %; 95% CI, 22%-33%; P<0.001); among the remote-from ischemic zone approaches, all (fastigial nucleus stimulation, subthalamic vasodilator area stimulation, and dorsal periaqueductal gray stimulation) reduced infarct volumes by approximately half. On formal rating scales, CHS studies had the lowest risk of bias, and CHS had the highest overall quality of intervention-level evidence supporting readiness to proceed to clinical testing. Conclusions- Electrical stimulation reduces final infarct volume across preclinical studies. CHS shows the most robust evidence and is potentially appropriate for progression to early-stage human clinical trial testing as a promising neuroprotective intervention.

Keywords: acute stroke; central nervous system; electrical stimulation; meta-analysis; neuroprotection.

Figure 1.

Forest plot shows the neuroprotective effect of electrical stimulation across multiple preclinical studies

Figure 2.

Shows an asymmetric Funnel plot likely due to in-between studies heterogeneity and over-representation of positive effects among smaller fastigial nucleus stimulation (FNS) studies. Due to some missing studies over the non-significant right lower area of the plot, the presence of publication bias was suggested, although was not statistically significant based on regression models (P=0.63 based on Egger’s and P= 0.45 based on Peters’ regression tests for bias).

Figure 3.

Study-level risk of bias ratings. Risk of bias items based on CAMARADES and STAIR recommendations^-, . For individual items: green indicates low risk of bias; yellow indicates some concerns; red indicates high risk of bias; white indicates unclear risk of bias. Total scores can range from 0 to 24, with scores of 16-24 indicating low risk of bias; scores of 8-15 indicating intermediate risk of bias; and scores of 0-7 indicating high risk of bias.

Figure 4.

Intervention-level evidence quality ratings and readiness-for-translation scoring. Quality items based on STAIR recommendations. Green indicates high evidence quality; Yellow indicates intermediate evidence quality; Red indicates low evidence quality. Note that for the two stimulation subtypes of anodal and electrical hemispheric stimulations, red was allotted to all the quality items due to lack of benefit of the two simulation strategies. Total scores can range from 0 to 18, with scores of 12-18 indicating high readiness-for-translation; scores of 6-11 indicating intermediate readiness-for-translation; and scores of 0-5 indicating low readiness-for-translation.