Efficacy and safety of remote ischemic conditioning for acute ischemic stroke: A comprehensive meta‐analysis from randomized controlled trials

Abstract Background and Purpose Remote ischemic conditioning (RIC) is a remote, transient, and noninvasive procedure providing temporary ischemia and reperfusion. However, there is no comprehensive literature investigating the efficacy and safety of RIC for the treatment of acute ischemic stroke. In the present study, we performed a comprehensive meta‐analysis of the available studies. Methods MEDLINE, Embase, the Cochrane Library database (CENTRAL), and ClinicalTrials.gov were searched before Sep 7, 2022. The data were analyzed using Review Manager 5.4.1 software, Stata version 16.0 software, and R 4.2.0 software. Odds ratio (OR), mean difference (MD), and corresponding 95% CIs were pooled using fixed‐effects meta‐analysis. Results We pooled 6392 patients from 17 randomized controlled trials. Chronic RIC could reduce the recurrence of ischemic stroke at the endpoints (OR 0.67, 95% CI [0.51, 0.87]). RIC could also improve the prognosis of patients at 90 days as assessed by mRS score (mRS 0–1: OR 1.29, 95% CI [1.09, 1.52]; mRS 0–2: OR 1.22, 95% CI [1.01, 1.48]) and at the endpoints assessed by NIHSS score (MD −0.99, 95% CI [−1.45, −0.53]). RIC would not cause additional adverse events such as death (p = 0.72), intracerebral hemorrhage events (p = 0.69), pneumonia (p = 0.75), and TIA (p = 0.24) but would inevitably cause RIC‐related adverse events (OR 26.79, 95% CI [12.08, 59.38]). Conclusions RIC could reduce the stroke recurrence and improve patients' prognosis. Intervention on bilateral upper limbs, 5 cycles, and a length of 50 min in each intervention might be an optimal protocol for RIC at present. RIC could be an effective therapy for patients not eligible for reperfusion therapy. RIC would not cause other adverse events except for relatively benign RIC‐related adverse events.


| INTRODUC TI ON
Stroke is the second leading cause of mortality and disability worldwide, accounting for 6.55 million deaths and 143 million disability-adjusted life years per year. 1 The most common type of stroke is the ischemic stroke (IS), which accounts for 87% of all strokes. 2 For these patients, the most intuitive means of treatment to restore blood flow before major brain damage occurs is to remove the blockage by a combined injection of clot-dissolving drugs or use of a mechanical device, or both. [3][4][5] Unfortunately, only less than 5% of stroke patients can benefit from these therapies, mainly due to the narrow therapeutic window and the risk of inducing ischemia-reperfusion injury (IRI), increasing lesion size and worsening blood barrier breakdown leading to cerebral edema and hemorrhage. [6][7][8][9][10] In view of this condition, an innovative and effective way to extend the therapeutic windows and to mitigate further brain injury is urgently needed. By targeting multiple molecular pathways associated with cell death, remote ischemic conditioning (RIC) is thought to remotely recruit neuroprotective pathways as a neuroprotective strategy. 11 RIC is an inexpensive, remote, noninvasive intervention that has been successfully demonstrated in several preclinical studies in the kidney, heart, and brain since its inception in 1986. [11][12][13] RIC provides temporary episodes of reversible ischemia through repeated inflations and deflations of a blood pressure limb cuff, with the intent of protecting remote organs such as the brain or heart from subsequent ischemic injury. 14 RIC can be divided into three types based on the initiation of conditioning: preconditioning (RIPreC, RIC applied before ischemia), perconditioning (RIPerC, RIC applied after the onset of ischemia and before reperfusion), and postconditioning (RIPostC, RIC applied during reperfusion). 15 Currently, the possible mechanisms of RIC include neurovascular protection by inducing anti-inflammatory effects and neuronal protection against excitotoxicity, coupled with mitochondrial protection, circulating inflammasome activation and/or transcriptional regulation of neuroprotective pathways, and enhancement of collateral circulation. [16][17][18][19] Although there have been previous systematic reviews and meta-analyses of RIC as a neuroprotective therapy, most of them focused on preclinical animal models. [20][21][22][23][24] However, translating treatment effects on stroke from animal ischemia models to clinical reality is often a great challenge due to nonphysiological experimental conditions in animals that do not reflect human conditions. 25 In addition, others made the summary of the completed RCTs on RIC in IS patients with a high percentage of participants lost to follow-up, a limited number of recruited subjects, and a lack of comprehensive analyses. [26][27][28][29] The standard for the optimal protocol and initiation time of RIC is important but still inconclusive. 30 Given that a systematic evaluation of the efficacy and safety of RIC in AIS and the optimal protocol of RIC has not yet been performed, we conducted a study to provide comprehensive evidence of the neuroprotective effects of RIC on patients with AIS through a meta-analysis.

| ME THODS
A meta-analysis was performed in conformity with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA , Table S1) guidelines. 31 We did not prepare a protocol and our study was not registered.

| Searchstrategy
We systematically searched MEDLINE, Embase, the Cochrane Library database (CENTRAL), and Clini calTr ials.gov for any peerreviewed research articles published before Sep 7, 2022. Search keywords included two aspects, including stroke and RIC, and search terms were adjusted for different databases. After removing duplicate and not directly relevant studies, two investigators (XJK and FW) screened each remaining article by reading the title, abstract, etc., to determine whether the research met the predefined inclusion criteria. Disagreements between the two investigators were resolved through sufficient discussion. In case of persistent divarication, a third investigator (ZYY) intervened to resolve the divarication. Detailed search strategies are provided in Table S2.

| Inclusion/Exclusioncriterion
Studies were included in our research if (a) the type of studies were randomized controlled trials (RCTs); (b) participants were adults with the onset of acute ischemic stroke based on a combination of clinical examination results, according to the diagnostic criteria for AIS used by researchers to enroll patients in their RCTs; (c) the RIC intervention consisted of RIPreC, RIPerC, and RIPostC, which were the same procedure but differed based on the protocol; (d) the control consisted of the sham group that required application of a blood pressure cuff or other occlusive devices (without complete blood flow occlusion such as cuff inflation to 30 mmHg) or blank group without sham procedure.
Studies were excluded if (a) the type of study was a single-arm trial, completed RCTs but without available results and RCTs with low compliance rate and non-RCTs; (b) participants were diagnosed with gradual-onset cerebral ischemia such as cerebral small-vessel disease and patients undergoing cardiovascular interventional surgery or any other type of cardiac surgery; (c) studies without outcomes of the present research.

| Outcomemeasures
The primary efficacy outcome was the recurrence of ischemic stroke at the endpoint. The secondary efficacy outcomes were based on the National Institutes of Health Stroke Scale (NIHSS, range: 0-42, with higher scores indicating more severe neurological impairment) scores, which is the most widely used measure of the severity of presenting stroke and a dominant predictor of patient functional outcome, 32 and modified Rankin Scale (mRS, range: 0-6, with higher scores indicating worse prognosis) scores, which is considered to be the standard clinical endpoints in acute stroke trials. 33 The secondary efficacy outcomes included the following: NIHSS score at the endpoint, excellent outcome (mRS 0-1) at 90 days, favorite outcome (mRS 0-2) at 90 days, and dependency (mRS 3-5) at 90 days. The safety outcomes were RICrelated adverse events including arm pain assessed by visual scale, redness or swelling of arms, skin petechiae on arms, dizziness or nausea and headache, death, intracerebral hemorrhage events including intracerebral hemorrhage and hemorrhagic transformation, pneumonia and TIA.

| Dataextraction
Two investigators (XJK and FW) independently extracted data from each included study and recorded the details on an extraction form.
The extracted data consisted of three aspects as follows: (a) baseline information of the included trials (first author, year, registration number, publication, country, total number of participants, type of disease), patient characteristics (age range, mean age, gender, details of receiving reperfusion therapy, NIHSS score), and outcome events; (e.g., mRS score at 90 days), the number of participants with the event in each group. All extracted data were thoroughly reviewed by the same two authors, and discrepancies were resolved by sufficient discussion with a third reviewer (ZYY). Where necessary, data were extracted from graphs using GetData Graph Digitizer software.

| Dataanalysis
The study data were analyzed using Review Manager 5. For continuous data, mean differences (MD) and their corresponding 95% confidence intervals (95% CI) were calculated using the inverse variance method with a fixed-effect model. For dichotomous data, odds ratios (ORs) and their corresponding 95% CIs were calculated with the Mantel-Haenszel method, also using a fixed-effect model.
Statistical heterogeneity between studies was assessed using Chisquared and I 2 statistics. The statistical test of heterogeneity was considered significant if the p-value was <0.10, and we defined a value greater than 50% as substantial heterogeneity. Where there was substantial heterogeneity, a heterogeneity test was performed using a Labbe plot, followed by a sensitivity analysis was conducted.
Publication bias was assessed using the Egger weighted regression statistic and visual inspection of contour-enhanced funnel plots. 34

| Subgroupanalysis
We performed subgroup analyses to examine heterogeneity by analyzing differences in the number of limbs occluded (unilateral or bilateral), the number of RIC cycles (4 or 5 cycles), the length of each RIC intervention (40 or 50 min), the type of RIC classified by the duration of the intervention (acute, delayed and chronic), and the time to initiate RIC (during ischemia before reperfusion therapy, during reperfusion after reperfusion therapy, during ischemia without reperfusion therapy). Worth mentioning, in the subgroup of type of RIC, 11 acute RIC is performed during the transfer of patients to comprehensive stroke centers, for use with reperfusion therapy (e.g., 1 session of RIC/day within 24 h of onset); chronic RIC is performed daily for several months for chronic neurological conditions such as intracranial atherosclerotic stenosis (ICAS) (e.g., 1 session of RIC/ day for no less than 180 days); and the duration of intervention of delayed RIC is between the first former two types (e.g., 1 session of RIC/day for more than 24 h to no more than two weeks). In addition, whether RIC was used as a primary or adjuvant therapy to reperfusion therapy was important, because the timing of therapy initiation has received considerable attention. 30 In the present study, patients who received reperfusion therapy accepted RIC therapy during ischemia or reperfusion, and those who did not receive reperfusion therapy accepted RIC during different phases of ischemia.

| Riskofbiasandqualityassessment
Two investigators (XJK and FW) independently assessed the risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions. 35 Disagreements were resolved by discussion or, if necessary, by consultation with a third investigator (ZYY). We assessed the risk of bias according to the following domains: selection bias, performance bias, detection bias, attrition bias, reporting bias, and other possible bias. For each domain, we classified each stud as "low risk," "unclear risk," or "high risk." We summarized the evidence by creating "Summary of findings" tables using GRADEproGDT (https://grade pro.org/), into which we imported data from RevMan 5.4.1. We used the GRADE approach to rate the quality of evidence as high, moderate, low, or very low, according to the following five considerations: study limitations, imprecision, consistency of effect, indirectness, and publication bias.

| Descriptionofincludedstudies
The study selection process has been shown in the flow diagram ( Figure 1). Using the search strategy described above, a total of 1299 records were retrieved from preselected databases by Sep 7, 2022.
After removing duplicates and records that were not directly relevant, 17 studies remained and were pooled in meta-analysis after applying the exclusion criteria. The primary information of the included studies is shown in Table 1 Table S3.

| Overalloutcomeofefficacyandsafety
We pooled 6392 patients from 17 randomized controlled trials. All the forest plots exported from RevMan 5.4.1 are shown in Figure S1.  for each outcome are shown in Figure 2. To further explore the efficacy and safety of RIC for patients diagnosed with AIS, subgroup analysis was performed and the results are summarized in Table S4.

| Subgroupanalysisforsecondary efficacyoutcomes
For NIHSS scores at the endpoints, RIC showed advantages over

| Riskofbiasassessmentand qualityoftheevidence
The risk of bias summary is shown in Figure S3 and the bias of each study is shown in Figure S4. Except for one study 40 whose randomization was based on the odevity of patient card IDs, which had a high risk of selection bias (random sequence generation), and three stud-  Tables S6 and S7. Except for the recurrence of ischemic stroke, excellent outcome (mRS 0-1), and favorite outcome (mRS 0-2), which were of high quality, the quality of the remaining outcomes was moderate. F I G U R E 2 Summary of the overall efficacy and safety of RIC on ischemic stroke. † NIHSS score at the endpoint was the only continuous outcome thus we did not map the forest plot in this figure. ‡ The effect size of RIC-related adverse events was far beyond the scale of this figure (0-2) thus we did not map the forest plot in this figure. The column of p-value, statistics for testing differences between RIC and control group within one factor (ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).

| DISCUSS ION
As shown in Figure 3A, our study showed that bilateral upper limbs applied in the RIC group could significantly improve the prognosis of patients which is different from the results of the previous preclinical meta-analysis. 20,21 It has been reported that one in four patients with IS has silent peripheral arterial disease 43 with a more rapid decline in lower limb-threatening ischemia, 44 for which it has been suggested that the upper limb would be the most appropriate to ensure safety. Most previous meta-analyses 20,21,24 showed it is effective to use both one and two limbs in RIC treatment, but they used young male rodents with a notable absence of animals with comor- The column of p-value, statistics for testing differences between RIC and control group within one factor (ns, not significant; *p < 0.05; **p < 0.01).
As shown in Figure 4, our analysis supported that chronic RIC might be beneficial. However, no statistical difference was found between acute and delayed RIC in reducing the recurrence of stroke.
According to our studies, it is suggested that chronic RIC is a good The column of p-value, statistics for testing differences between RIC and control group within one factor (ns, not significant; *p < 0.05; **p < 0.01).

F I G U R E 5
Summary of the subgroup analysis results for the time to initiate RIC. † NIHSS score at the endpoint was the only continuous outcome thus we did not map the forest plot in this figure. ‡ The effect size of RIC-related adverse events was far beyond the scale of this figure (0-2) thus we did not map the forest plot in this figure. The column of p-value, statistics for testing differences between RIC and control group within one factor (ns, not significant; *p < 0.05; ***p < 0.001; ****p < 0.0001).

| CON CLUS ION
Overall, chronic RIC could significantly reduce the recurrence of stroke at the endpoint, and improve patients' prognosis at 90 days or the endpoints assessed by mRS score and NIHSS score. Furthermore, intervention on bilateral upper limbs, 5 cycles, and a length of 50 min in each intervention might be an optimal protocol for RIC currently. For patients ineligible for reperfusion therapy, RIC might be an effective way to recover from damage. Inspiringly, RIC might not cause other adverse events besides relatively benign RIC-related adverse events.

AUTH O RCO NTR I B UTI O N S
XK involved in drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; and analysis or interpretation of data. ZY

CO N FLI C TO FI NTE R E S TS TATE M E NT
The authors declare no conflicts of interest.

DATAAVA I L A B I L I T YS TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.