Remote ischaemic conditioning for stroke: unanswered questions and future directions

Dose and frequency

The minimally effective and most effective dose and frequency of RIC is unclear. With most studies employing between 3–5 × 5 minute cycles of upper limb ischaemia at 200 mm Hg once daily, a single ‘dose’ of RIC can last between 30 and 50 mins. Table 1 illustrates the variation in RIC protocols used in ongoing clinical trials. By identifying the dose–response relationship between RIC and any potential clinical benefits clinicians will be better able to counsel patients with shorter cycle lengths potentially resulting in improved concordance with therapy.

In addition to the RESIST trial, investigating single and repeated RIC in acute stroke, two further studies are investigating the optimal dose and frequency of chronic RIC (NCT03105141 and ChiCTR1800014403). In the first study, NCT03105141, 600 patients with symptomatic intracranial stenosis and a recent stroke or high-risk TIA are stratified into 1 of 10 treatment groups. All the treatment groups receive automated RIC daily for 1 year with substudies assessing the effect of different cuff pressures (200 mm Hg or 40 mm Hg above systolic), cycle length (4, 5 or 6 mins), cycle number (4, 5 or 6) and frequency (once daily vs twice daily) on stroke recurrence. At the same centre, ChiCTR1800014403 will randomise 273 individuals with symptomatic intracranial stenosis to either 3×3 mins or 5×5 mins cycles of RIC twice daily and compare the effect of a shorter RIC intervention on secondary prevention. Timing of RIC in Acute Stroke: The optimal timing of RIC delivery in acute stroke is unclear. Among the major trials of RIC in acute stroke there is significant heterogeneity in patient inclusion criteria and the timing of RIC (table 2). For instance, in RICAMIS (NCT03740971) only acute moderate-severe ischaemic stroke (National Institute of Health Stroke Scale, NIHSS 6–16) cases confirmed on neuroimaging and within 48 hours of symptom onset are included while, in the RESIST study (NCT03481777), RIC is commenced in a prehospital phase for patients with a presumed stroke (based on the combined Cincinnati Prehospital Stroke Scale and Postural Assessment Scale for Stroke) within 4 hours of symptom onset. A schematic of the timing and duration of RIC in the larger acute stroke trials is illustrated in figure 2. If the beneficial effects of RIC in acute stroke require early intervention, then adequately powered studies of RIC delivered in the prehospital phase are essential to establish its efficacy.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

A schematic of RIC timing and duration in major trials of acute stroke. RIC, remote ischaemic conditioning.

Duration: It remains unclear whether there is a ceiling effect of repeated, daily RIC; if RIC is still efficacious when delivered on a less frequent basis, this may increase uptake and concordance. In chronic stroke, RIC has been used daily for up to 300 days.15 Meng et al reported on a proof-of-concept study of twice daily bilateral upper limb RIC for 300 days in patients with symptomatic intracranial stenosis. Those in the treated arm (n=38) experienced significant reductions in stroke recurrence after 90 and 300 days of RIC compared with controls treated with medical therapy alone (n=30).15 The Kaplan-Meier analysis of stroke recurrence indicates deviation of RIC group from controls after around 60 consecutive days of RIC, however, it is unclear whether a shorter duration of chronic RIC (eg, 60 days) would be sufficient to sustain a clinically beneficial effect over the long term. The maximum follow-up time in current RIC studies is 12 months, which may be suboptimal given the follow-up times required to prove effectiveness in secondary prevention trials of antiplatelets and statins.31 32 Taken together, this could help establish an agreed protocol of either a single course of RIC, periodic RIC or lifelong RIC. Both the use of RIC for stroke prevention, potentially lifelong, and long term trial follow-up, are significant investments for patients and trialists, respectively. The discovery of validated biomarkers for vascular protection could shorten follow-up times and identify patients responding to treatment, negating the need for unnecessary long term treatment.

Unilateral versus bilateral RIC

While bilateral RIC may theoretically produce a larger ischaemic stimulus, no study is currently comparing the clinical impact of unilateral vs bilateral RIC. Unilateral RIC may be more convenient for individuals with hemiparesis and may allow concurrent completion of daily tasks during therapy. Furthermore, in the context of hyperacute stroke, the use of unilateral RIC enables the other arm to be used for venous access and intravenous thrombolysis.

For unilateral RIC, it is unclear if the marginal gain from repeated conditioning in the same limb diminishes over time. As such, it would be helpful to establish whether alternating the site of delivery has a role in maintaining the conditioning stimulus over time. Furthermore, it is unclear whether the use of unilateral RIC in the paretic arm yields the same conditioning response as RIC in the non-paretic arm; experimentally, nerve transection has been shown to block the effect of preconditioning,33 however, this has not been established for organic central lesions.

Upper limb versus lower limb RIC

While lower limb RIC has the theoretical potential to induce a larger ischaemic stimulus, a study of RIC in healthy volunteers showed no significant differences in the production of nitrite, a potential signalling mediator of the conditioning response, after upper limb or lower limb RIC in healthy volunteers.34 However, as discussed, given the conditioning response likely involves multiple signalling cascades that occur over different time frames, there is insufficient evidence to conclude that upper limb and lower limb RIC are equipotent. We found only five studies investigating lower limb RIC (RESCUE-BRAIN (NCT02189928),29 PreLIMBS (NCT02411266), RIPC-SAH (NCT02381522), NCT04039399, NCT04038697). The preponderance of interventions using upper limb RIC may reflect increased practicality for both patients and carers. A qualitative analysis of patient experience with both upper and lower limb RIC may reveal key barriers to longer-term compliance that can help investigators optimise intervention strategies.

Safety and tolerability

Multiple clinical trials suggest that RIC is safe,20 21 even when performed twice daily on both arms for 300 days,15 including in the elderly.16 However, relatively few studies of chronic RIC have reported on compliance, therefore, it is difficult to ascertain whether stroke patients are able to carry out RIC for sustained periods of time. In Meng et al,15 51 patients with symptomatic intracranial stenosis were randomised to receive twice daily bilateral arm conditioning for 300 consecutive days but only 38 (74.5%) completed the study. Long-term data on larger cohorts will help inform clinicians on likely compliance rates.

A small number of side effects have been reported including skin petechiae,16 but no studies have indicated increased risks of venous thrombosis or rhabdomyolysis. It is unclear if chronic RIC affects sensory or motor function of the conditioned limbs; studies that include these detailed clinical and electrophysiological assessments as outcome measures should help fill this knowledge gap.

A consistent exclusion criterion for several clinical trials of RIC in stroke has been known peripheral vascular disease or venous thromboembolism affecting the conditioned limb. While there is a theoretical risk of harm in these patients, the fact that prior studies have not highlighted incident risk of these complications, and that studies are investigating upper limb RIC in patients with known lower limb peripheral arterial disease,35 should provide some reassurance to clinicians.

RIC and ET

The usefulness of RIC alongside ET for large vessel occlusion (LVO) has attracted much research interest. Numerous questions arise. The first is whether RIC can be delivered parallel to ET without causing significant treatment delays. The second is whether, RIC performed alongside recanalisation of a culprit vessel may influence cerebral autoregulation in a manner that increases the risk of reperfusion injury or haemorrhagic transformation. Hypertension and impaired cerebral autoregulation often already accompany acute ischaemic stroke.36 37 The third is whether the effect of general anaesthetic diminishes the effect of RIC; as previous studies have indicated, the use of propofol inhibits the ischaemic conditioning response.38

Manual versus automated RIC

Automated RIC devices are widely available. While RIC has been marketed as a potentially low-cost, non-invasive intervention for vascular diseases, the costs of obtaining multiple devices, cuff and battery replacements, may prove to be prohibitively expensive in low and middle income countries where the rates of cerebrovascular disease remain high.39 As such, studies which deploy manual RIC (with sphygmomanometers and digital timers) will help develop the global translatability of RIC.

Three studies plan to compare the effect of RIC in those undergoing ET (REVISE-2 - NCT03045055, PROTECT-I - NCT03915782, RICE-PAC - NCT03152799) (online supplemental material). As well as revealing efficacy, these will highlight the practical challenges of delivering RIC to patients in a time-sensitive manner.

Target populations

Intracranial stenosis: The majority of studies of chronic RIC have been performed in individuals with significant intracranial stenosis. This is an important population to study for two reasons: high stroke recurrence rates in this population mean a lower sample size is needed to detect a significant treatment effects, and they are a subgroup where significant improvements in care can potentially be made. However, while intracranial atherosclerosis contributes to ischaemic stroke in a large proportion of individuals globally, it is a less common mechanism of stroke in European and North American populations where extracranial carotid disease, cardioembolic stroke and small vessel disease are more common.40 41 As such, studies need to extend to individuals without intracranial stenosis to establish the utility of RIC in the wider setting of secondary stroke prevention. Furthermore, most studies of chronic RIC in ischaemic stroke have been performed in East Asia; while there is no identifiable reason why the response to RIC would be different in a different population, it is important for similar studies to be undertaken outside of Asia to identify whether the chronic RIC is both tolerated and beneficial in a wider global setting.

The influence of comorbidities and medications: It is recognised that the impact of RIC may diminish with age.42 However, Meng et al,16 demonstrated that RIC may reduce stroke recurrence in those over 80 thereby suggesting that ageing alone may not suppress the beneficial effects of RIC.

Diabetes mellitus is a common risk factor for ischaemic stroke, however, evidence from both animal models and clinical studies suggest that diabetes, particularly diabetic neuropathy, may abolish the protective effect of RIC.43 For instance, a recent study of chronic, daily RIC in patients with diabetes with peripheral arterial disease failed to show any positive impact on vascular or neuronal function35; potentially linked to impairment of the neurogenic pathways thought to potentiate the action of RIC.43 The theoretical risk of compounding nerve damage with repeated limb compression in chronic RIC may also be a reason for excluding such patients from trials, and their ultimate under-representation in the literature. Subgroup analyses of larger randomised controlled trials or pooled datasets may help determine if the presence of diabetes truly impacts on the clinical effects of RIC.

Animal models also suggest that the conditioning response is diminished by the presence of comorbidities such as hypercholesterolaemia, hypertension and obesity.42 However, this has not yet been confirmed in clinical studies of RIC in stroke. Similarly, several medications that are commonly prescribed in individuals with cardiovascular disease including nitrates, atorvastatin and nicorandil may activate similar intracellular signalling cascades to RIC.44 However, clinical data are lacking with respect to understanding whether this would potentiate or abrogate the effects of RIC in patients taking these medications.

Cardioembolic stroke

Studies of RIC in ischaemic stroke typically exclude individuals with known cardioembolism from causes such as atrial fibrillation or flutter. In these patients, stroke often results from the sudden migration of thrombus from the heart to the cerebral arteries, a pathway that RIC may not obviously affect at first glance. However, the impact of RIC on secondary prevention of cardioembolic stroke is of interest for several reasons. First, the seminal studies investigating the protective effect of ischaemic preconditioning were performed in animal models where ischaemia was precipitated acutely,22 23 a mode of infarction more akin to cardio-embolism than atherosclerotic disease. It may be that slow atheromatous progression and intracranial stenosis leads to a degree of endogenous pre-conditioning that reduces the marginal impact of RIC compared with those with cardio-embolic stroke. Consistent with this, previous studies have indicated that the volume of ischaemia resulting from cardioembolism is larger than that due to a similar sized clot from the carotid/intracranial circulation.45 Furthermore, the most common cause of cardio-embolism, atrial fibrillation, develops predominantly in later life and in the context of cardiovascular risk factors such as hypertension and diabetes mellitus. As such, many patients with atrial fibrillation have associated endothelial dysfunction,46 another pathway potentially impacted by RIC.47

Small vessel ischaemic disease

Small vessel ischaemic disease (SVID) is a common comorbidity and complication associated with stroke and is a precursor for the development of vascular cognitive impairment (VCI) and vascular dementia. The role of RIC in SVID has attracted interest due its potential effects on CBF and inflammatory markers. In SVID, hypoperfusion due to microangiopathy can precipitate neurodegeneration through disruption of the blood–brain barrier and neuroinflammation.48Declines in cognitive function have also been associated with reduced microvascular perfusion.49 Small pilot studies of patients with established SVID or those with VCI have shown that RIC applied twice daily (5×5 min cycles) for between 6 and 12 months can result in improvements in CBF velocities, reductions volume of white matter lesions, and variable improvements in domains of cognitive function.50–52 Investigations are currently under way to evaluate whether once daily application, for a shorter duration of 1 month has similar effects with greater treatment adherence.53

RIC and exercise

Experimental and clinical evidence suggests shared biological mechanisms for the effects of RIC and exercise, including increased heat shock proteins and increased endothelial nitric oxide signalling.54 It is unclear whether there is any synergistic effect of RIC when performed in addition to regular exercise, and no current studies evaluating this in chronic stroke. While exercise confers many physiological benefits in stroke patients,55 numerous barriers to exercise poststroke exist including depression, fatigue and physical disability. As such, if there are physiological effects of RIC that mirror exercise, then this may be a useful long-term ancillary treatment when patients are physically unable to exercise.

Underlying mechanism and biomarkers

Applying an external pressure to occlude blood flow to a limb initiates transmission of a signal that mediates the RIC response. Humoral, neurogenic and immune mediated pathways have been proposed as transmission and end organ effector mechanisms.9 After the transmission of the conditioning stimulus, an array of possible biochemical and physiological changes may contribute to clinically meaningful outcomes in stroke patients including increased CBF,15 angiogenesis and increased collateral circulation,27 antiplatelet effects,56 57 synaptogenesis,58 immune regulation, mitochondrial function and autophagy regulation,47 (figure 3). The latency of these biological mediators and effectors may vary and are likely to underpin differences in effects of acute and chronic RIC.9 Establishing the underlying mechanisms and final effectors of RIC will prove essential in optimising intervention development, trial design and establishing its efficacy and safety in wider clinical practice. Furthermore, by characterising the intracellular pathways associated with the beneficial effects of RIC, appropriate drug targets may be identified, leading the way towards pharmacological conditioning.59

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Proposed mechanism of action of remote ischaemic conditioning in stroke.

For subarachnoid haemorrhage, the putative biological mechanism for any potential benefit of RIC are not well characterised but may involve reducing vasospasm.60 In intracerebral haemorrhage, there are pathophysiological consequences that are shared with ischaemic stroke including endothelial dysfunction, impaired cerebral autoregulation, a proinflammatory mileu and mitochondrial dysfunction61; these have all been proposed as therapeutic targets of RIC in ischaemic stroke, and therefore, may prove to be of significance in intracerebral haemorrhage.

Circulating biomarkers

Previous studies have indicated potential serum signalling mediators of RIC including heat shock proteins, tumor necrosis factor (TNF)-α and interleukin 6 (IL-6).9 However, no established circulating biomarkers have been shown to be reliably correlate with the degree of conditioning response or clinical outcomes. Identifying a serum biomarker would be advantageous as it could offer a minimally invasive measure to: (1) establish which patients are responding to the treatment and (2) optimise the dose, frequency and duration of RIC treatment on an individualised basis. Several ongoing studies are investigating candidate circulating biomarkers including NCT02323425 (VEGF and bFGF), RESIST - NCT03481777 (eryNOS3 phosphorylation, microRNA and extracellular vesicle profile), NCT03546309 (S100-A4, MMP-9, bFGF, PDGF, VEGF) and NCT03335111 (IL-4, IL-6, TNF) (online supplemental material). Future studies of RIC in stroke should include focused analyses of candidate biomarkers, identified from experimental animal and human models, at different time points before, during and after both acute and chronic RIC. It is also important to validate which biomarkers are associated with the transmission of the ischaemic signal and which are associated with the biological effector response and clinical outcomes.

Radiological correlates

Animal models,27 and previous clinical studies,15 indicate that RIC can increase CBF as well as increasing the dynamic range of cerebral autoregulation in healthy volunteers.37 In the setting of acute ischaemic stroke, RIPerC may increase CBF (directly or via collaterals) to salvage the ischaemic penumbra.27 Several studies are investigating the effect of RIPerC on infarct volume and clinical outcomes acutely (online supplemental material), however, few trials use non-invasive measures such as Transcranial Doppler (TCD) to characterise the impact of RIC on blood vessel diameter and flow. Several studies investigating chronic RIC are measuring cerebral and skeletal muscle blood flow using a range of modalities including TCD, Arterial Spin Labelling and MR angiography (NCT03635177, NCT02323425 and NCT04038697). If increased cerebral or skeletal muscle blood flow is associated with positive clinical outcomes in the larger studies of chronic RIC then non-invasive assessments of blood flow could be used to establish response to RIC on a personalised level.

Economic analysis

The costs of automated conditioning devices, device maintenance and safety monitoring need to be factored into a cost–benefit analysis of RIC in stroke. Furthermore, additional care staff may be required to administer RIC to individuals who are unable to independently apply the device.

Recommendations for future trial design

Given the multitude of ongoing studies for RIC in stroke, a collaborative effort is required to synthesise the array of results into meaningful outcomes for stroke patients. Figure 4 represents a proposed schema for future clinical trial design, identifying six key areas which future studies should focus their attention in order to address the remaining uncertainties of RIC in stroke.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Optimising remote ischaemic conditioning (RIC) trial design. AIS, Acute Ischaemic Stroke; ET, endovascular therapy; ICH, Intracerebral haemorrhage; SAH, Subarachnoid Haemorrhage; tPA, Tissue Plasminogen Sctivator.