Benefits of GLP-1 Receptor Agonists for Stroke Reduction in Type 2 Diabetes: A Call to Action for Neurologists
Abstract
People living with diabetes are at higher risk for stroke and have a poorer prognosis following a stroke event than those without diabetes. Data from cardiovascular outcome trials and meta-analyses indicate that GLP-1RAs (glucagon-like peptide 1 receptor agonists) reduce the risk of stroke in individuals with type 2 diabetes. Accordingly, many guidelines now recommend the addition of GLP-1RAs to ongoing antihyperglycemic regimens to lower the risk of stroke in type 2 diabetes. The current work summarizes evidence supporting the use of GLP-1RAs for stroke reduction in people with type 2 diabetes and offers 2 new resources for neurologists who are considering GLP-1RAs for their patients—a list of frequently asked questions with evidence-based answers on safely initiating and managing GLP-1RAs, and a practical decision-making algorithm to assist in using GLP-1RAs as part of a stroke reduction strategy.
Glycemia, Type 2 Diabetes, and Stroke
Diabetes is associated with premature vascular aging and is a strong risk factor for stroke. Approximately 30% of patients presenting with stroke have known diabetes,1 and more are later determined to have a new prediabetes or type 2 diabetes (T2DM) diagnosis.2,3 Independent of other factors, the risk of ischemic stroke is approximately twice that in people with diabetes compared with those without.3 Risk factors for atherosclerotic disease such as hypertension and dyslipidemia are often comorbid in diabetes1 and further increase the risk of stroke.
Stroke outcomes are worse in individuals with diabetes. In the acute/subacute phase, hospitalized patients with both diabetes and stroke are more likely to suffer from complications such as urinary tract infection and pneumonia and have higher mortality rates.4 Long-term, people with both diabetes and stroke continue to be at higher risk for adverse events, including recurrent transient ischemic attack (TIA)/stroke, heart failure, and death.5
Despite improvements in screening for diabetes and an expanded diabetes and stroke reduction armamentarium, the prevalence of stroke in individuals with diabetes has remained relatively stable,6 underscoring a persistent care gap. Reducing stroke risk in people with diabetes has traditionally focused on lowering hemoglobin A1c (HbA1c; despite limited evidence for macrovascular benefit), management of comorbid modifiable vascular risk factors, and use of antithrombotic agents for those who have had a previous ischemic stroke or TIA.
Current stroke prevention guidelines7–9 recommend the use of GLP-1RAs (glucagon-like peptide 1 receptor agonists), a newer class of antihyperglycemic agents that reduce the risk of stroke in people with T2DM. The addition of the GLP-1RAs to the neurology toolkit offers an original and exciting opportunity to obtain a meaningful impact on stroke.
The current work stems from an expert forum that was convened in September 2021 to undertake a state-of-the-art review of the evidence supporting GLP-1RAs for stroke reduction in people with T2DM. Notably, it includes 2 novel resources for neurologists who may be considering GLP-1RAs for their patients—first, a list of frequently asked questions with evidence-based answers on how to safely initiate and manage GLP-1RAs; second a practical decision-making algorithm to assist in using GLP-1RAs as part of a stroke reduction strategy.
Interventions for Stroke Reduction in Type 2 Diabetes
What Is the Evidence for Nonglycemic Interventions in Stroke?
Modifiable behavioural changes that address exercise, diet, smoking cessation, and alcohol consumption remain the foundational therapies for people with T2DM.7,8,10 Equally important nonglycemic interventions for primary and secondary prevention of stroke in T2DM include lowering blood pressure to <130/80 mm Hg, managing lipid levels, and implementing antithrombotic strategies as necessary.7,8,10–12 The effect that nonglycemic agents have on stroke reduction in a diabetes setting is summarized in Table 1. Depending on the cause of the event, individuals who have experienced a TIA/stroke may require additional/alternate treatments to lower the risk of recurrent stroke such as anticoagulation for those with a cardioembolic source and consideration of revascularization for significant symptomatic carotid stenosis.7,8
Is There a Role for Intensive Glycemic Control in Stroke?
Despite the association between hyperglycemia and worse outcomes after acute ischemic stroke, aggressive treatment of hyperglycemia with insulin therapy following hyperacute or acute stroke does not appear to improve clinical outcomes.20 In the nonacute setting, meta-analyses indicate that intensive glycemic control with mainly non-GLP-1RA antihyperglycemic agents in T2DM can lower the risk of microvascular and certain macrovascular complications.21 The risk of nonfatal myocardial infarction (MI) and the composite of cardiovascular death, nonfatal MI, and nonfatal stroke were reduced, whereas the risk of stroke on its own was not.9,21 Conversely, a post hoc analysis of the landmark Steno-2 study (Intensified Multifactorial Intervention in Patients With Type 2 Diabetes and Microalbuminuria) reported that an aggressive multi-risk factor approach significantly reduced micro- and macro-vascular complications, including time to first stroke event (hazard ratio [HR], 0.31 [95% CI, 0.14–0.69]; P=0.004), compared with conventional care.22 Thus, while achieving lower glycemic levels on its own does not appear to reduce stroke risk, when used as part of a multifactorial approach, many complications of T2DM, including stroke, can be reduced with long-lasting effects.
Is There a Role for Non-GLP-1RA Antihyperglycemic Agents in Stroke?
In a meta-analysis of 48 randomized trials with 8 classes of antihyperglycemic therapies that reported stroke outcomes (Figure S1), statistically significant reductions in stroke were only found with the GLP-1RAs and thiazolidinediones.23 SGLT2is (sodium-glucose-transporter-2 inhibitors) reduce cardiorenal outcomes including the composite of major adverse cardiovascular events (MACE), hospitalization for heart failure and kidney end points across a spectrum of people with T2DM, but not stroke.23,24 Notably, although the SGLT2is do not appear to reduce stroke, people with T2DM and stroke were included in the SGLT2i CV outcome trials (CVOTs) and that subgroup experienced similar cardiorenal benefits as the rest of the populations.25,26 Accordingly, the current Canadian Stroke Best Practice Recommendations for the secondary prevention of stroke includes SGLT2is as a consideration for those with both stroke and T2DM to reduce other major cardiovascular outcomes.8
Thiazolidinediones increase insulin sensitivity and impact adipocyte differentiation. A meta-analysis of 5 thiazolidinedione studies that included stroke outcomes demonstrated a relative risk of 0.82 (95% CI, 0.69–0.98).23 However, only 1 of the 5 studies was performed in a population of only those with T2DM. The PROactive (Prospective Pioglitazone Clinical Trial in Macrovascular Events) study randomized 5238 people with T2DM and macrovascular disease to receive either pioglitazone or placebo in addition to standard-of-care. After a mean follow-up of 34.5 months, there was no difference in the primary 6-point composite end point (HR, 0.90 [95% CI, 0.80–1.02]; P=0.095).27 The main secondary composite end point of nonfatal MI, nonfatal stroke and all-cause death was reduced with pioglitazone (HR, 0.84 [95% CI, 0.72–0.98]; P=0.027) but the individual outcome of stroke was not.27 In a subgroup analysis of PROactive looking at the subgroup with stroke at baseline, pioglitazone did reduce fatal/nonfatal stroke (HR, 0.53 [95% CI, 0.34–0.85]; P=0.0085).28 The Spanish Society of Neurology’s Stroke Study Group accordingly states that pioglitazone may be added to standard diabetes therapy to reduce stroke recurrence in those with T2DM and prior stroke.9 However, among those without a prior stroke history, thiazolidinediones are not recommended for stroke prevention in T2DM.9
Is There a Role for GLP-1 Receptor Agonists in Stroke?
Emerging evidence has suggested that GLP-1RAs reduce the incidence of stroke in people with T2DM. GLP-1RAs have been studied in 8 large CVOTs in T2DM cohorts with a primary outcome of MACE, including cardiovascular death, nonfatal MI and nonfatal stroke.29–36 In the REWIND trial (Researching Cardiovascular Events With a Weekly Incretin in Diabetes) with dulaglutide and the SUSTAIN-6 trial (Trial to Evaluate Cardiovascular and Other Long-Term Outcomes With Semaglutide in Subjects With Type 2 Diabetes) with subcutaneous semaglutide, there were significant decreases in the primary MACE outcome and significant reductions in the prespecified secondary outcome of nonfatal stroke.29,30
Dulaglutide and Stroke Reduction
REWIND, powered for superiority, randomized 9901 people with T2DM and either cardiovascular disease (CVD) or cardiovascular risk factors to either dulaglutide 1.5 mg subcutaneous once weekly (QW) or placebo and followed them for a median of 5.4 years.30 The primary MACE outcome was reduced by 12% (HR, 0.88 [95% CI, 0.79–0.99]; P=0.026) with dulaglutide, while the nonfatal stroke component of MACE was reduced by 24% (HR, 0.76 [95% CI, 0.61–0.95]; P=0.017; Figure 1A). Although fatal/nonfatal stroke was significantly reduced by 24% and ischemic stroke by 25%, the decrease in fatal stroke or hemorrhagic stroke did not achieve statistical significance.37 Disabling stroke, defined as a modified Rankin Scale score of ≥3, was reduced by 26%, but in those who had a stroke, there was no difference in the distribution of stroke severity scores between dulaglutide and placebo.37 The impact of dulaglutide on stroke not differ according to sex, age, history of CVD, HbA1c, systolic blood pressure, or use of statins or renin-angiotensin-aldosterone system inhibitors. The stroke risk benefits were also notably independent of a history of prior stroke/TIA (Pinteraction=0.83).37 In Canada, dulaglutide is indicated as an adjunct to diet, exercise, and standard-of-care therapy to lower the risk of nonfatal stroke in adults with T2DM who have multiple cardiovascular risk factors or established CVD.38 In the United States, dulaglutide is indicated to reduce the risk of MACE in adults with T2DM who have established CVD or multiple cardiovascular risk factors.39
Semaglutide and Stroke Reduction
SUSTAIN-6, powered for noninferiority, randomized 3297 individuals with T2DM and either established CVD or ≥1 cardiovascular risk factor to semaglutide 0.5 mg, semaglutide 1.0 mg or placebo (all QW subcutaneous) and followed them for a median of 2.1 years.29 MACE was reduced by 26% (HR, 0.74 [95% CI, 0.58–0.95]; P<0.001 for noninferiority; P=0.02 for superiority, not prespecified) with semaglutide and the nonfatal stroke component of MACE was reduced by 39% (HR, 0.61 [95% CI, 0.38–0.99]; P=0.04; Figure 1B). Both doses of semaglutide yielded similar risk reductions for MACE and nonfatal stroke. There was no significant decrease in the outcomes of fatal/nonfatal stroke or fatal stroke alone, and only 5 study participants experienced a fatal stroke.29,40,41 The effect of semaglutide on nonfatal stroke did not differ according to the participants’ age, sex, history of CVD or importantly, MI/stroke history (Pinteraction=0.75).42 A post hoc pooled analysis of semaglutide in the SUSTAIN-6 trial and PIONEER 6 trial (Peptide Innovation for Early Diabetes Treatment) demonstrated a 32% reduction (HR, 0.68 [95% CI, 0.46–1.00]; P=0.048) in fatal/nonfatal stroke.43 The ongoing SOUL (A Heart Disease Study of Semaglutide in Patients With Type 2 Diabetes)44 and ASCEND PLUS45 trials should offer further insights on the impact oral semaglutide has on MACE, including stroke, in T2DM. Semaglutide is indicated in the United States to reduce the risk of MACE in adults with T2DM who have established CVD.46 There is currently no data for GLP-1RAs and stroke reduction in people without diabetes. However, the ongoing SELECT (Semaglutide Effects on Heart Disease and Stroke in Patients With Overweight or Obesity) trial (URL: https://www.clinicaltrials.gov; Unique identifier: NCT03574597) is evaluating the impact of semaglutide 2.4 mg QW in individuals with obesity or overweight with CVD but not diabetes and has nonfatal stroke as a prespecified secondary outcome.
Meta-Analyses of GLP-1 Receptor Agonists and Stroke
A meta-analysis of the 8 completed CVOTs with GLP-1RAs revealed statistically significant reductions in the combined outcome of fatal/nonfatal stroke versus placebo (HR, 0.83 [95% CI, 0.76–0.92]) as well as decreases in MACE, cardiovascular death, fatal/nonfatal MI, and hospitalization for heart failure.13 The specific outcomes of nonfatal stroke alone or fatal stroke alone were not reported. In another meta-analysis of 8 CVOTs, nonfatal stroke was reduced by 16% (HR, 0.84 [95% CI, 0.76–0.94]), but the outcomes of fatal/nonfatal stroke and fatal stroke were not reported.47 Earlier studies reported similar declines in fatal/nonfatal stroke in meta-analyses of 4 to 7 CVOTs, decreases in nonfatal stroke across 7 CVOTs, but no reduction in fatal stroke when meta-analyzing 5 or 6 studies.9,23,48–51 The exact reasons for the neutral outcomes in some studies are unknown although it is plausible that they resulted from the smaller sample sizes, fewer events, and low power to detect a benefit. Table S1 summarizes the details of these meta-analyses. We performed a meta-analysis that included all 8 CVOTs using the Cochrane Collaboration’s Review Manager Software Package (RevMan 5.3) and found significant decreases versus placebo in fatal/nonfatal stroke (HR, 0.83 [95% CI, 0.76–0.92]), nonfatal stroke (HR, 0.84 [95% CI, 0.76–0.93]), but no reduction in fatal stroke (HR, 0.80 [95% CI, 0.63–1.03]; Figure 2). Given that there were only 221 fatal stroke outcomes across the 8 trials, it is likely that this outcome is underpowered for demonstrating superiority. The significant 17% relative risk reduction in fatal/nonfatal stroke in the meta-analysis of 8 CVOTs with GLP-1RAs is numerically similar to many of the proven interventions for stroke reduction in diabetes that were described earlier (ie, acetylsalicylic acid,14 statins,15 blood pressure lowering, ramipril,11 low-dose rivaroxaban,17 icosapent ethyl, and evolocumab19; Table 1).
There are no head-to-head randomized trials with cardiovascular outcomes comparing GLP-1RAs to SGLT2 inhibitors. An indirect comparison via network meta-analysis of 176 trials showed that GLP-1RAs compared with SGLT2is reduced nonfatal stroke by 17%,52 but data from a real-world cohort of 557 109 individuals suggested that SGLT2is and GLP-1RAs provide comparable protection against atherosclerotic CVD in people with T2DM.53
GLP-1 Receptor Agonists and Recommendations for Stroke Reduction
Diabetes Canada recommends the use of GLP-1RAs to provide cardiovascular protection in those with CVD or risk factors.54 In the same vein, the 2020 Canadian Stroke Best Practices update for secondary prevention of stroke,8 2021 Spanish Society of Neurology consensus statement,9 2021 American Heart Association/American Stroke Association Guideline for the Prevention of Stroke in Patients with Stroke and Transient Ischemic Attack,7 and the Diabetes, Cardiorenal, and Metabolic Diseases Multispecialty Task Force55 all endorse GLP-1RAs for reducing stroke risk in individuals with T2DM. Specifically, the Canadian Stroke Best Practices recommends GLP-1RAs for patients with stroke and T2DM who have not met their glycemic targets with standard antihyperglycemic agents8; the American Heart Association/American Stroke Association guideline suggests GLP-1RAs as second-line to metformin regardless of baseline HbA1c in individuals with established atherosclerotic CVD to avoid further vascular events7; the Diabetes, Cardiorenal, and Metabolic task force endorses the use of a long-acting GLP-1RAs for primary and secondary prevention of stroke/TIA in diabetes55; but the Spanish guidelines only recommend GLP-1RAs for prevention of fatal and nonfatal stroke in individuals with T2DM and either established vascular disease or high vascular risk.9
Possible Mechanisms for Stroke Reduction With GLP-1 Receptor Agonists
Several mechanisms are believed to be responsible for the cardiovascular-metabolic-renal benefits of GLP-1RA.56 The stroke reduction advantages are thought to arise from a combination of increased neuroprotection and decreased cerebral atherosclerosis (Figure 3).57–59 At the molecular and cellular levels, GLP-1RAs diminish oxidative stress, reduce apoptosis, and limit the formation and accumulation of advanced glycation end-products, all of which help attenuate the proinflammatory components of cerebral atherosclerosis. GLP-1RA therapy also directly improves endothelial function, promotes angiogenesis, cerebral blood flow, and neurogenesis alongside decreasing neuronal damage as well as amyloid plaque build-up and infarct volume within the cerebral circulation. GLP-1RAs further lower stroke by indirectly reducing blood pressure, promoting weight loss, and enhancing HbA1c control. In a mediation analysis, HbA1c lowering accounted for 54% of the stroke reduction in REWIND.37 Furthermore, meta-regression analysis of GLP-1RA CVOTs demonstrates a linear relationship between HbA1c reduction and stroke reduction.60,61
Algorithm for GLP-1 Receptor Agonist Use for Stroke Prevention in Type 2 Diabetes
Given the available evidence supporting the use of GLP-1RAs to reduce the risk of stroke among those living with T2D and the corresponding recommendations from various guidelines, we believe it is important for neurologists to consider GLP-RAs in appropriate patients. While we concur with the Spanish Society of Neurology that a secondary prevention study following participants with T2DM and a stroke history is warranted to provide more robust data on the utility of GLP-1RAs for prevention of stroke recurrence in T2DM,9 we deem the lack of heterogeneity for stroke reduction in REWIND and SUSTAIN-6 for the subgroups with prior stroke/TIA and prior MI/stroke respectively as sufficient evidence to recommend dulaglutide or semaglutide for stroke reduction in those with prior stroke. This aligns with the recommendations of other groups.7,8,55 Figure 4 includes a practical algorithm to aid in GLP-1RA initiation, and Table 2 provides answers to some frequently asked questions with recommendations on how to safely address common concerns around the use of GLP-1RAs.
Briefly, we suggest that a GLP-1RA with proven stroke reduction data (dulaglutide 1.5 mg QW or semaglutide 0.5/1.0 mg QW) be considered, independent of baseline HbA1c, for those with T2DM who have experienced a prior stroke/TIA or are at high-risk for stroke. Factors such as cost/coverage, side effects, need for weight loss, injection route, and priority for hypoglycemia reduction, should also be discussed with the patient before prescribing a GLP-1RA. Any dipeptidyl peptidase-4 inhibitor (eg, alogliptin, linagliptin, saxagliptin, sitagliptin, and vildagliptin) should be stopped upon initiation of a GLP-1RA. If the HbA1c is <8%, sulfonylureas (eg, glibenclamide, glimepiride, glipizide, glyburide, and tolbutamide) and insulin will require downward adjustments to lessen the risk of hypoglycemia.
We are cognizant that the opportunities for neurologists to initiate GLP-1RAs may vary according to practice styles and underscore that GLP-1RAs can be safely started at any point during the patient journey. GLP-1RAs may be initiated during hospital admission for acute stroke or upon discharge from hospital after an acute stroke. While some neurologists may suggest in a discharge summary that the primary care provider initiate the prescription, others may elect to start a GLP-1RA as part of a treatment plan during an outpatient stroke prevention clinic visit or may prefer to refer to a diabetes care provider instead.
Conclusion and a Call to Action
Prevention of stroke is important in the clinical management of T2DM. Studies of the GLP-1RAs dulaglutide and subcutaneous semaglutide, as well as GLP-1RA meta-analyses, demonstrate a reduction in the rates of stroke. Future trials of GLP-1RAs with stroke as a primary outcome would be beneficial regarding the evidence base for GLP-1RAs and stroke reduction. Based on current evidence, we recommend a call to action for neurologists and their utilization of GLP-1RAs for patients with T2DM at high risk of stroke as part of their stroke prevention strategies.
Acknowledgments
Dr Verma holds a Tier 1 Canada Research Chair in Cardiovascular Surgery (University of Toronto). Editorial support, under the direction of the authors, was provided by Hwee Teoh, PhD (HTaq Biomedical Editorial and Education Services Inc). Graphic support (Figure 3) was provided by Cassie Hillock-Watling (Cassie HW Visuals).
Source: https://www.ahajournals.org/doi/full/10.1161/STROKEAHA.121.038151
