American Heart Association

Higher Admission Heart Rate Associated with Death and Poor Functional Outcome in ICH

Alexander E. Merkler, MD

Qiu M, Sato S, Zheng D, Wang X, Carcel C, Hirakawa Y, et al. Admission Heart Rate Predicts Poor Outcomes in Acute Intracerebral Hemorrhage: The Intensive Blood Pressure Reduction in Acute Cerebral Hemorrhage Trial Studies. Stroke. 2016

Intracerebral hemorrhage is a devastating disease with a one-month mortality of 40%. Larger ICH volume, older age, and hematoma expansion are some of the factors associated with both poor functional outcome and death. Admission heart rate (HR) has previously been shown to predict higher mortality in coronary artery disease and ischemic stroke, but its impact on patients with ICH is unknown. 


In this study, Drs. Qiu et al perform a post-hoc analysis on data from the INTERACT trials to evaluate the effect of admission heart rate on outcome in ICH. Clinical outcomes included mortality and functional outcome (mRS) at 90 days. Imaging outcome was hematoma growth on 24 hour CT. HR was divided into quartiles (<65, 65-74, 75-84, ≥85) and Cox logistic regression was used to account for possible confounders in the relationship between admission HR and the outcomes of interest.

Of 3185 patients with ICH, patients with higher admission HR had higher BP, more frequent intraventricular extension of hematoma, and were less likely enrolled in China; patients with lower admission HR were more likely older, more often were taking a beta-blocker or antithrombotic, and had larger hematomas that were less likely to be in a lobar location.  Overall, higher admission HR was associated with higher mortality (adjusted hazard ratio for HR ≥85 vs. <65: 1.5; CI, 1.07-2.11). In addition, higher admission HR was associated with worse functional outcome at 90 days (adjusted odds ratio 1.33; CI 1.08-1.63). There was no significant association between admission HR and hematoma expansion on 24 hour CT.

Similar to coronary artery disease and ischemic stroke, admission HR appears to be associated with increased mortality and poor functional outcomes in patients with ICH. As the authors suggest, perhaps higher admission HR is a marker of poor general health, dehydration, anemia, or a marker of cardiac disease, all of which are predictors of poor outcome after stroke. One major limitation is the lack of adjustment for heart rate variability, which has also been shown to be associated with poor outcomes after stroke.  

In conclusion, admission HR is associated with increased mortality and poor functional outcome in patients with ICH. 

Elevated Blood Pressure Significantly Associated with Risk of Vascular Dementia

Danny R. Rose, Jr., MD


Vascular dementia is the second most common cause of dementia, but many aspects of the disease are poorly understood. In particular, there is conflicting evidence regarding the relationship between blood pressure and vascular dementia. Elevated blood pressure in midlife has been found to have a positive association with future development of dementia, but several other studies have found low blood pressure in old age to be associated with an increased risk of dementia. One possible explanation of these findings is that it represents “reverse causality,” meaning vascular dementia is responsible for low blood pressure by decreasing sympathetic drive. Blood pressure medication may also play a confounding role in this association. Emdin et al. sought to further clarify this association by conducting an analysis of 4.28 million individuals without vascular disease or dementia, supplemented with an analysis of a prospective population-based cohort of patients with TIA and stroke.

The study included patients from age 30 to 90 and excluded patients with pre-existing cardiovascular disease to minimize the potential of reverse causality related to advanced age and cardiovascular disease causing reduced blood pressure, respectively. The endpoint of the primary analysis was an inclusive definition of vascular dementia based on ICD 10 coding and was inclusive of patients with co-existing Alzheimer’s disease. Secondary analysis excluded these patients and excluded patients treated with medications commonly used to treat AD. The first four years of follow-up were excluded in the primary analysis to mitigate the effect of patients with undiagnosed dementia. Cox models, stratified by practice, were used to determine hazard ratios for the association for clustering of patients by practice. The primary analysis was adjusted for age, sex, body mass index and smoking status. The Oxford Vascular Study cohort was used to confirm findings independently.

Out of a cohort of 4.28 million individuals free of vascular disease and dementia, 14,934 cases were reported to have vascular dementia. After excluding for presentations during the first four years of follow-up, 11,114 cases were included. The association between usual SBP and risk of vascular dementia followed a linear progression within the age groups of 30-50 and 51-70. The age group of 71-90 did not show a significant association. The strength of association decreased with increasing age category. Overall for individuals aged 70 years or less at baseline, 20 mm Hg higher usual SBP was associated with a 26% higher risk of vascular dementia (HR 1.26 CI 1.17, 1.34). Significant negative associations with systolic and diastolic blood pressures were observed for the age group 71-90, but after excluding for the first eight years of follow-up, no significant association was observed. Adjusting for patients in the primary care cohort that had TIA and stroke events reduced the HR to 1.18, indicating that 30% of the excess risk of vascular dementia per 20mm Hg higher SBP is mediated through risk of future stroke and TIA. The OXVASC cohort did not show a relationship between the most recent SBP or DBP in patients relative to their diagnosis of new dementia, but did show significant positive associations with DBP and SBP in 5-9 years prior to the TIA/stroke and particularly 10-20 years prior.

This study supports prior positive associations between blood pressure in mid-life and vascular dementia and also suggests that elevated blood pressure attributes a significant risk for the development of vascular dementia at least until the age of 70. The authors’ rationale for excluding confounders in the cohort appears to be sound and had the intended effect of strengthening the associations studied. The study refutes previous reports of a negative association with blood pressure and vascular dementia in the elderly, likely in part due to the aforementioned adjustments, strengthening the authors’ hypothesis of reverse causality. This study represents by far the largest analysis of the association between blood pressure and risk of vascular dementia and although it is susceptible to limitations related to the diagnosis of dementia in a primary care setting, it represents a significant advancement in our understanding of the complex pathophysiology of vascular dementia.


Blood Biomarkers in Ischemic Stroke

Ilana Spokoyny, MD


Blood biomarkers are an active research interest, with the potential for predicting ischemic stroke via identification of novel risk pathways. The authors of this paper tested the associations between ischemic stroke and three blood markers: procalcitonin (PCT), copeptin, and midregional-pro-atrial natriuretic peptide (MRproANP). These biomarkers were chosen because they represent three different pathophysiological processes. Procalcitonin is associated with bacterial infections and was chosen with the hope of finding a link between infection and ischemic (especially non-cardioembolic) stroke. Copeptin is a hypothalamic stress hormone, which was chosen because chronic activation of the hypothalamic-pituitary-adrenal axis and the sympathetic nervous system may promote vascular risk factors. MRproANP was hypothesized to be a marker of hemodynamic dysfunction and thereby a potential marker of (especially cardioembolic) stroke.


A nested case-control study was performed among initially stroke-free patients enrolled in the Northern Manhattan Study (NOMAS). 172 cases of first ischemic stroke were compared to 344 randomly selected controls. Primary endpoint (occurrence of first ischemic stroke) was detected using phone calls and prospective monitoring of hospital admissions and discharges. Patients with intracerebral hemorrhage were excluded.

PCT levels in the blood may reflect ongoing subclinical inflammatory processes triggered by bacterial endotoxins. It is unclear why there was an association of PCT with specifically small vessel stroke subtype. MRproANP was associated with cardioembolic stroke subtype, which may represent cardiac pathology (which has also been linked to ANP) causing cardioembolic strokes. Interventions based on these associations must be validated in prospective clinical trials, but this is a strong first step in developing better risk models and blood-based diagnostic tools for ischemic stroke. If these are validated and incorporated into clinical practice, one could imagine using MRproANP to help distinguish cardioembolic from non-cardioembolic sources of cryptogenic stroke, or using procalcitonin as one of the factors helping to distinguish mild strokes or TIAs from mimics. 
The adjusted hazard ratios for new ischemic stroke comparing the highest to lowest quartiles of PCT and MRproANP were 1.9 (1.0-3.8) and 3.5 (1.6-7.5), respectively. Stroke etiology was determined using based on the TOAST criteria. High PCT levels (top quartile) were associated with small vessel stroke (HR 5.1, 1.4-18.7) and high MRproANP levels (top quartile) were associated with cardioembolic stroke (HR 16.3, 3.7-70.9). 


Relevance of Carotid Plaque Characteristics for Ischemic Stroke and Coronary Heart Disease


Citing the systemic nature of atherosclerosis, the authors of this paper studied the association between extracranial carotid atherosclerosis features and prior ischemic stroke (IS) and coronary heart disease (CHD). 

Patients were selected from the Rotterdam Study. Participants with at least 2.5mm of carotid intima media thickness (IMT) were eligible for this study. Of 3,795 eligible participants, 1,982 underwent MRI of bilateral carotid arteries. MRI was performed on 1.5-Tesla scanners with a carotid artery protocol. All plaques of at least 2mm of thickness were assessed for intraplaque hemorrhage (IPH), lipid core, and calcification. Additionally, wall thickness and degree of stenosis were determined. Participants’ history was queried for prior IS and CHD (non-fatal MI or myocardial revascularization). Covariates were age, sex, smoking status, lipid measurements, BMI, diabetes, and hypertension.

Binomial logistic regression was used to determine the association between individual plaque characteristics and history of IS and CHD. Men and women were examined separately.

One thousand seven hundred thirty-one participants were ultimately included. The mean age was 73 years, and 55% were male. A majority of patients had a history of smoking. IS and CHD were much more common in men, as were vascular risk factors. The mean carotid wall thickness was 3.6mm, and the mean degree of stenosis was 13%. IPH was present in 35%, lipid core in 41%, and calcification in 82%. In terms of outcomes, 105 had a history of IS, and 199 had a history of CHD.

In the overall population, in multivariate analysis, only degree of stenosis was associated with IS. For men, stenosis and IPH were associated with IS. In both the overall population and when stratified by sex, only carotid stenosis was associated with CHD. 

The key finding is that whereas plaque thickness or stenosis were associated with both IS and CHD, IPH was only associated with IS (in men). The dependence on prevalence data and the lack of clinical data (e.g. stroke laterality or mechanism) limit the conclusions that can be drawn from this study. However, this work may foreshadow and stimulate prospective, mechanistically enlightening studies that capitalize on the granularity of MRI data.

Intracranial Atherosclerosis and Coronary Atherosclerosis: Two Twigs from the Same Vascular Branch

Peggy Nguyen, MD

Chung J-W, Bang OY, Lee MJ, Hwang J, Cha J, Choi J-H, et al. Echoing Plaque Activity of the Coronary and Intracranial Arteries in Patients With Stroke. Stroke. 2016

Atherosclerosis is a diffuse process that can affect both the coronary and carotid arteries, but while previous studies have suggested a strong correlation between coronary atherosclerosis and extracranial carotid atherosclerosis, the correlation with intracranial atherosclerosis is less clear. Whereas the mechanism of myocardial infarction from coronary atherosclerosis is likely more similar to ischemic stroke caused by extracranial atherosclerosis, ischemic stroke caused by intracranial atherosclerosis typically falls into two etiologies: branch occlusive disease-type (B-type), where atherosclerosis occludes a perforating artery, versus coronary-type plaque rupture of plaque (C-type), where the atherosclerotic plaque ruptures, causing a shower of multiple embolic infarcts distally. This study attempts to characterize intracranial plaque phenotypes and correlate asymptomatic coronary artery disease (CAD) with intracranial atherosclerotic disease (ICAD) burden.


A total of 81 patients were included the final analysis, drawn from a population of patients admitted within 7 days of symptom onset for treatment of acute ischemic stroke with intracranial atherosclerosis. Patients who had known histories of coronary artery disease were excluded. B-type ICAS was differentiated from C-type ICAS in both anterior and posterior territory strokes. An ICAD score was calculated on the basis of intracranial atherosclerotic burden, with 0 points given for stenosis less than 50%, 1 point for stenosis of 50-99% and 2 points for an occlusion, with all involved intracranial vessels summed for a total score

Asymptomatic CAD was quite common, with a prevalence of just over 80% in the study population. The prevalence of asymptomatic CAD was relatively similar in both B-type and C-type ICAS groups (48% vs 52%) and, as might be expected, the burden of ICAD was positively correlated with the burden of CAD, although non-calcified coronary artery plaque morphology was independently associated with C-type ICAS. As non-calcified coronary plaque increased, remodeling also increased in the symptomatic arteries of patients with ICAS.

This study provides evidence of a positive relationship between coronary and intracranial atherosclerotic burden, and that coronary artery plaque composition (calcified vs non-calcified) might predict intracranial atherosclerosis morphology. The investigators suggest that this should prompt us as clinician to take a more holistic approach to the entire vascular system, rather than solely focus on, for example, the cerebral vasculature, or the coronary arteries. Certainly this might prompt the clinician to, when faced with a stroke patient with C-type ICAS, be more cognizant of the type of likely associated CAD burden, but a study evaluating whether this might also be predictive of acute coronary syndrome, would be of additional benefit.

PCI Setting Has Little Effect on Post-Stroke Cardiovascular Outcomes and Mortality

Benjamin Kummer, MD

Myint PK, Kwok CS, Roffe C, Kontopantelis E, Zaman A, Berry C, et al. Determinants and Outcomes of Stroke Following Percutaneous Coronary Intervention by Indication. Stroke. 2016.

Percutaneous coronary intervention (PCI) is infrequently complicated by stroke, which tends to occur more frequently after emergent PCI, and is highly morbid when it occurs. However, the relationship between PCI setting (emergent vs. elective) or stroke subtype (ischemic vs. hemorrhagic) and stroke risk factors, as well as post-PCI stroke outcomes, remains unclear. Myint et al. sought to address these gaps in knowledge by studying approximately 560,000 PCI patients from the British Cardiovascular Intervention Society database, using in-hospital major adverse cardiovascular events (MACE) and 30-day mortality as outcomes.

The rate of any post-PCI stroke was low (0.13%), 80% of strokes were ischemic, and any post-PCI stroke was associated with increased odds of poor outcome irrespective of setting (ORs, no stroke as reference: elective MACE 21.05, p<0.001, emergent MACE 6.25, p<0.001, elective mortality 37.90, p<0.001, emergent mortality 5.00, p<0.001). ORs were higher in the elective vs. emergent setting. The odds of mortality after hemorrhagic stroke (ORs: elective 175.24, p<0.001, emergent 21.50, p<0.001) were higher than after ischemic stroke irrespective of setting (ORs: elective 17.61, p<0.001, emergent 3.11, p<0.001), consistent with the high mortality associated with ICH.

Generally, setting did not significantly affect the odds of poor outcome—except for ischemic stroke after emergent PCI, which had lower odds of in-hospital MACE than after ischemic stroke in the elective setting (OR 0.44, 95%CI 0.22-0.86). The authors attributed this to higher usage of “advanced” anti-platelet medications / GP IIb/IIIa inhibitors in emergent PCI settings.

Stroke risk factors differed by PCI setting. Female sex and cardiovascular comorbidities such as CABG, cardiogenic shock, ventilator/circulatory support and STEMI were associated with increased odds of stroke in the emergent setting, whereas female sex and the use of GP IIb/IIIa inhibitors were the main risk factors for the elective setting. There were also differences in risk factors according to stroke subtype; thrombectomy was associated with ischemic stroke, whereas thrombolysis was associated with hemorrhagic stroke. The main limitations were unclear ascertainment of stroke, absent functional status data and post-stroke management, plus the impossibility of gauging the timing of stroke in relation to PCI.

Overall, the study confirms the rarity and known high morbidity and mortality of post-PCI stroke once it occurs, and also shows that risk factors vary between stroke subtypes and settings. However, PCI setting does not seem to have major impact on the likelihood of poor outcome after post-PCI stroke.

Residual Arterial Stenosis after Endovascular Thrombectomy: a Relationship with in Situ Thrombo-occlusion and Reocclusion Rates

Mark R. Etherton, MD, PhD
The advent of efficacious endovascular thrombectomy (EVT) for ischemic stroke secondary to acute occlusion of proximal anterior circulation vessels has allowed for the characterization of occlusive lesions. Understanding the underlying pathology of these occlusive lesions could be informative for predicting the success of the endovascular intervention as well as prognostication of clinical outcomes.

In the present study, Hwang et al. characterized residual stenosis post EVT at the site of the arterial occlusive lesion in an Asian population with acute ischemic stroke secondary to middle cerebral artery M1 occlusion. In this population with a high prevalence of intracranial atherosclerotic disease (ICAD), the authors’ hypothesis was that residual stenosis, as defined by the Arterial Occlusive Lesion (AOL) scale, would be sequelae of in situ thrombo-occlusion (IST) with underlying ICAD. Angiographic imaging during EVT and follow up imaging (MR or CT angiography) 5 to 7 days post-procedure was performed to assess stenosis.
Out of 163 patients enrolled in the study, 74 patients (45.5%) had partial recanalization (AOL 2) on post-procedural angiography. Rates of favorable clinical outcomes at 3 months (defined as mRS of 2 or less) did not differ between the group with partial (AOL 2) versus complete recanalization (AOL 3). Forty patients (24.5%) in the study were determined to have IST as their stroke etiology, and all of these patients had residual stenosis present on the post-procedural angiogram (AOL 2). 27% of patients with partial recanalization compared to only 1.1% of patients with complete recanalization developed instant reocclusion during EVT. In addition, those patients with partial recanalization during EVT were more likely to have worse stenosis or occlusion (10.8% vs 1.1%) on follow-up imaging. On multivariable regression analysis, delayed reocclusion in patients with partial recanalization was predicted by excellent baseline collateral-flow (OR 8.477; 95%CI 1.169-61.464) and neurological worsening post-procedure (OR 10.388; 95%CI 1.287-83.876).

It is important to note that the authors used a radiologic-based approach to identify IST that was based on the presence and severity (>50%) of residual stenosis and the presence of ICAD. The majority of patients with IST determined using these criteria were classified as partial recanalization with residual stenosis exceeding 50% on follow-up angiography. This study suggests that IST is a common cause of large vessel occlusion in an Asian population and that residual stenosis is associated with increased risk of reocclusion and early neurologic deterioration. Going forward, work should include controlling for endovascular approach utilized and characterizing the pathologic correlates for radiographic-determined IST.

Inverse Relationship between HDL2-C Subfraction and Carotid Intima-Media Thickness

Alexander E. Merkler, MD 

Increased carotid-intima media thickness (cIMT) is associated with future cerebrovascular events. Although previous data has shown an inverse relationship between high density lipoprotein cholesterol (HDL-C) and future cerebrovascular events, this association has been challenged in recent trials; therefore, it is uncertain whether HDL-C subfractions have differential effects on cerebrovascular risk.

Using the Northern Manhattan Study (NOMAS), Dr. Tiozzo et al. evaluate the relationship between HDL-C subfractions and cIMT. NOMAS is a prospective cohort study designed to determine stroke incidence and risk factors in a multiethnic urban population in Manhattan.

Nine hundred eighty-eight stroke-free participants with available data on HDL-C subfractions and cIMT measurements using high-resolution ultrasound were evaluated. HDL-C was assessed as HDL2-C and HDL3-C subfractions, and total HDL. cIMT was calculated as the composite measure of the near and the far wall of IMT in the common, internal, and bifurcation of the carotid artery on both sides of the neck. After controlling for demographics, vascular risk factors, LDL, and triglyceride levels, both HDL2-C and total HDL-C were inversely associated with cIMT; the association was even more robust among patients with diabetes. No association was found between HDL3-C and cIMT.

The main limitation is the lack of temporality between HDL-C subfraction measurement and cIMT assessment.

Overall, the results suggest an inverse relationship between total HDL and HDL2-C and cIMT, but not between HDL3-C and cIMT. If confirmed, these results may lead to HDL subfraction targeted therapies to reduce the risk of cerebrovascular disease.

Fractional Anisotropy Change in Acute Phase of Stroke and its Correlation with Motor Recovery at 3 Months

Qing Hao, MD, PhD


“Doctor, will my father move his right arm again? How much do you think he can recover from this stroke?” As stroke neurologists, we are often asked about the prognosis after the stroke and most time the answer would be, “I am not exactly sure”.
Neuroimaging, especially MRI brain, has been very helpful in prognostication. Previous studies performed on chronic stroke patients demonstrated integrity/ atrophy of cortical spinal tract (CST) and signals indicative Wallerian degeneration (WD) on MRI in the chronic phase closely correlated with motor outcome.

Any neuroimaging markers in the acute phase of stroke can predict motor outcome? CST lesion load (based on initial motor impairment, lesion size and location) and CST integrity still played important roles. When talking about CST integrity, we can’t ignore fractional anisotropy (FA), an imaging marker derived from diffusion tensor imaging, which quantifies the organization (e.g. degree of alignment) and integrity of white matter tracts.
In this article, the author focused on FA in acute phase—they sought to investigate if FA difference in acute phase can be detected and how it predict motor outcome measured by upper extremity Fugl-Meyer score at 3 months.

Retrospective analyses were performed on a prospectively collected cohort of 58 patients with first time ischemic hemispheric stroke. MRI was done within 80 hours after stroke onset. FA values were determined in two regions of interest for 50 patients: cerebral peduncle and a stretch of the CST caudal to each stroke lesion (Nearest-5-Slice – N5S).
The authors were able to detect subtle asymmetry of FA changes (lower FA in the ipsilesional CST), most significantly in the slice that was closest to the ischemic lesion, not in the cerebral peduncle. The slope of the FA laterality index for the nearest-5-slices showed a weak but significant prediction (R2=0.11, p=0,022) for 3-month UE-FM score in univariate analysis, not in multivariate analysis. Not surprisingly, initial UE-FM, weighted CTS lesion load and days of therapy were stronger predictors (R2 =0.69, 0.71 and 0.249 respectively, p< 0.001) for 3-months UE-FM score.
CST integrity may be a dynamic change and its predictive value for motor recovery may be different at various stages after stroke. We look forward to more precise prediction models that would help us answer those challenging prognostication questions.

No Benefit for Poststroke Dysphagia with Pharyngeal Electrical Stimulation: a Randomized Controlled Trial

Mark R. Etherton, MD, PhD

Occurring in up to 50% of patients, post stroke dysphagia represents a common byproduct of stroke. The long term clinical and rehabilitation consequences of post stroke dysphagia are widespread as dysphagia increases the risk of aspiration and subsequent pneumonia, and malnutrition. At present, there is no definitive treatment for dysphagia aside from behavioral techniques like modified diet consistencies or surgical intervention with gastrotomy tubes.

Pharyngeal electrical stimulation (PES), is a developing technique that has shown promise in small phase II trials for the treatment of post stroke dysphagia. A meta-analysis of three trials showed that PES reduced aspiration and dysphagia while also being safe. Based on these findings, the authors in this study performed a large, randomized-controlled phase III trial of PES in patients with subacute poststroke dysphagia.

One hundred sixty-two patients with recent ischemic or hemorrhagic stroke and evidence of dysphagia, as defined by a penetration aspiration score (PAS) of 3 or greater on video fluoroscopy, were randomized to PES versus sham treatment daily for 3 days. Primary outcome measure was radiological aspiration as measured by PAS at 2 weeks. Secondary outcomes included PAS at 12 weeks, functional outcomes (mRS, BI, and NIHSS at 12 weeks) and clinical dysphagia scores.

With PES treatment, there was no difference in the PAS at 2 weeks (sham 3.6 +/- 1.9; PES 3.7 +/- 2.1; p=0.6) or 12 weeks (sham 3.0 +/- 2.1; PES 3.3 +/- 2.2; p=0.41). The authors also found no effect with PES on the secondary outcomes involving clinical dysphagia and functional outcomes. PES was well tolerated, as there was no difference in adverse or device-related events.

In summary, while PES was safe, it did not improve dysphagia in this randomized controlled trial of patients with post stroke dysphagia. Important consideration for this study and its interpretations, however, include high intraindividual variability in PAS and the authors’ note that up to 58% of patients may have received subtherapeutic stimulation. Based on these findings in conjunction with the positive prior meta-analysis, it would seem that further studies with different outcome measures are warranted.