Prediction Of Recurrent Stroke Research Articles

Spotlight on the December 7 Issue

#### Notable in Neurology This Week This issue features an article that investigates the frequency of neurologic manifestations in patients with coronavirus disease 2019 (COVID-19); another determines the frequency of autoimmune encephalitis associated with SARS-CoV-2. A report from the American Academy of Neurology Guideline Subcommittee provides a practice advisory update summary for stopping antiseizure medication in patients who have been seizure-free for several years. ### Carotid Plaque Inflammation Imaged by PET and Prediction of Recurrent Stroke at 5 Years ![Graphic][1]</img> In a pooled analysis of 3 prospective … [1]: /embed/inline-graphic-1.gif

Prediction of recurrent stroke among ischemic stroke patients with atrial fibrillation: Development and validation of a risk score model.

There is currently no validated risk prediction model for recurrent events among patients with acute ischemic stroke (AIS) and atrial fibrillation (AF). Considering that the application of conventional risk scores has contextual limitations, new strategies are needed to develop such a model. Here, we set out to develop and validate a comprehensive risk prediction model for stroke recurrence in AIS patients with AF. AIS patients with AF were collected from multicenter registries in South Korea and Japan. A developmental dataset was constructed with 5648 registered cases from both countries for the period 2011‒2014. An external validation dataset was also created, consisting of Korean AIS subjects with AF registered between 2015 and 2018. Event outcomes were collected during 1 year after the index stroke. A multivariable prediction model was developed using the Fine-Gray subdistribution hazard model with non-stroke mortality as a competing risk. The model incorporated 21 clinical variables and was further validated, calibrated, and revised using the external validation dataset. The developmental dataset consisted of 4483 Korean and 1165 Japanese patients (mean age, 74.3 ± 10.2 years; male 53%); 338 patients (6%) had recurrent stroke and 903 (16%) died. The clinical profiles of the external validation set (n = 3668) were comparable to those of the developmental dataset. The c-statistics of the final model was 0.68 (95% confidence interval, 0.66 ‒0.71). The developed prediction model did not show better discriminative ability for predicting stroke recurrence than the conventional risk prediction tools (CHADS2, CHA2DS2-VASc, and ATRIA). Neither conventional risk stratification tools nor our newly developed comprehensive prediction model using available clinical factors seemed to be suitable for identifying patients at high risk of recurrent ischemic stroke among AIS patients with AF in this modern direct oral anticoagulant era. Detailed individual information, including imaging, may be warranted to build a more robust and precise risk prediction model for stroke survivors with AF.

Carotid Plaque Inflammation Imaged by PET and Prediction of Recurrent Stroke at 5 Years.

To determine whether carotid plaque inflammation identified by 18F-fluorodeoxyglucose (18FDG)-PET is associated with late (5-year) recurrent stroke. We did an individual-participant data pooled analysis of 3 prospective studies with near-identical study methods. Eligible patients had recent nonsevere (modified Rankin Scale score ≤3) ischemic stroke/TIA and ipsilateral carotid stenosis (50%-99%). Participants underwent carotid 18FDG-PET/CT angiography ≤14 days after recruitment. 18FDG uptake was expressed as maximum standardized uptake value (SUVmax) in the axial single hottest slice of symptomatic plaque. We calculated the previously validated Symptomatic Carotid Atheroma Inflammation Lumen-Stenosis (SCAIL) score, which incorporates a measure of stenosis severity and 18FDG uptake. The primary outcome was 5-year recurrent ipsilateral ischemic stroke after PET imaging. Of 183 eligible patients, 181 patients completed follow-up (98.9%). The median duration of follow-up was 4.9 years (interquartile range 3.3-6.4 years, cumulative follow-up period 901.8 patient-years). After PET imaging, 17 patients had a recurrent ipsilateral ischemic strokes at 5 years (recurrence rate 9.4%, 95% confidence interval [CI] 5.6%-14.6%). Baseline plaque SUVmax independently predicted 5-year ipsilateral recurrent stroke after adjustment for age, sex, carotid revascularization, stenosis severity, NIH Stroke Scale score, and diabetes mellitus (adjusted hazard ratio [HR] 1.98, 95% CI 1.10-3.56, p = 0.02 per 1-g/mL increase in SUVmax). On multivariable Cox regression, SCAIL score predicted 5-year ipsilateral stroke (adjusted HR 2.73 per 1-point increase, 95% CI 1.52-4.90, p = 0.001). Plaque inflammation-related 18FDG uptake improved identification of 5-year recurrent ipsilateral ischemic stroke. Addition of plaque inflammation to current selection strategies may target patients most likely to have late and early benefit from carotid revascularization. This study provides Class I evidence that in individuals with recent ischemic stroke/TIA and ipsilateral carotid stenosis, carotid plaque inflammation-related 18FDG uptake on PET/CT angiography was associated with 5-year recurrent ipsilateral stroke.

Right to left shunting detection by contrast-enhanced transcranial color-coded duplex among patients with cryptogenic stroke

BackgroundContrast-enhanced transcranial duplex (c-TCD) might be more sensitive than transesophageal echo (TEE) for detection of right to left shunting (RLS), which misses some cases with substantial RLS and might be valuable for prediction of recurrent stroke or transient ischemic attack in patients with PFO. Our aim is to detect sensitivity and specificity of contrast-enhanced TCD in detection of RLS among stroke patients with patent foramen ovale (PFO) in comparison to TEE.MethodsTEE and contrast-enhanced TCD for cryptogenic stroke patients with PFO were done to detect right to left shunting.ResultsOn testing characteristics of TCD in detecting RLS compared to the gold standard of TEE, TCD sensitivity was 85.7%, specificity was 100%, negative predictive value was 96.55%, and positive predictive value was 100%.ConclusionWe concluded that PFO is considered an important hidden etiology for ischemic stroke. Contrast-enhanced TCD is considered sensitive method for detection of right to left shunting among PFO patients.

PRIMARY AND RECCURENT STROKES: CLINICAL AND PATHOGENETIC RELATIONS

Based on the data of clinical and laboratory-instrumental study of 81 patients with recurrent stroke, (including 63 cases after the primary ischemic stroke and 18 cases after the primary hemorrhagic stroke), we compared clinical-pathogenetic correlations between primary and recurrent strokes. A trend towards the localization of a focus of recurrent stroke in the same hemisphere in case of the similar to the primary stroke clinical form and in the other hemisphere in case of changing the clinical form (ischemic to hemorrhagic and vice versa) was found. Among the pathogenetic types of ischemic stroke, lacunar and hemoreologic types were most stable in recurrent stroke compared to the primary one. In other variants, the pathogenesis of recurrent stroke changed more often. The authors concluded that further research in this direction would help to solve issues in prediction of recurrent stroke and improvement of secondary prevention.

Improved Prediction of Recurrent Stroke and Detection of Small Volume Stroke

Improved Prediction of Recurrent Stroke and Detection of Small Volume Stroke

Abstract WP530: A Machine Learning Approach in Prediction of Recurrent Stroke

Introduction: Although recurrent stroke is clinically important, prediction of recurrent stroke is less well known. Machine learning encompasses a body of statistical approaches that can detect complex interaction patterns from multi-dimensional data. ML is gradually being adopted in medical science. The purpose of this study was to illustrate the possibility of using machine learning as an aid in prediction of recurrent stroke. Methods: We obtained clinical data from a consecutively registered, hospital-based acute stroke registry. Using a dataset of 4,766 acute ischemic stroke patients, predictive models were built using 1) linear 2) non-linear/non-parametric machine learning algorithms. The accuracy obtained was compared to that of the baseline model built without the clinical data. Results: There were 810 patients with recurrent stroke after ischemic stroke or transient ischemic attack (median follow-up, 365 days). The performances of the various machine learning models were compared. The accuracy of the baseline model was 31.4%, but the best non-linear model showed an accuracy of 64.6%, which was significantly higher than that of non-informative random guessing (p=0.002). Overall, the models using the nonlinear algorithms showed better accuracy than the linear ones. Conclusions: The high performance of the developed models demonstrated the predictive capacity of recurrent stroke and justified the application of machine learning algorithms. However, the practical application of machine learning models may need refinement and larger-scale data collection.

Abstract WP166: Prediction of Recurrent Stroke in Lacunar Stroke Patients: Performance of the Small Vessel Disease (SVD) Score in the SPS3 Trial

Introduction: The previously published cerebral SVD Score is a four-point rating scale based on the four cardinal MRI markers of SVD: moderate-severe white matter hyperintensities, , lacune(s), moderate-severe enlarged perivascular spaces (PVS), and microbleed(s). We evaluated the score’s ability to predict recurrent stroke in the Secondary Prevention of Small Subcortical Strokes (SPS3) Trial. Methods: One point each was awarded for these 4 MRI findings: max periventricular Fazekas score = 3 and/or max subcortical Fazekas score &gt; 2; &gt;1 cerebral microbleed 2 (i.e. &gt; 11 PVS unilaterally); &gt; 1 old lacune (3-15 mm) on FLAIR/T1. Annualized rates of recurrent stroke were computed assuming a Poisson model, and c-statistics were calculated for the SVD score model and for two other previously published SPS3 derived models. Results: Of 3020 participants, 1137 had complete data available for SVD scoring. Prevalence of SVD scores 0, 1, 2, 3, and 4 were 19% (n=219), 29% (n=325), 24% (n=277), 18% (n=200), and 10% (n=116). PVS (55%) were most common, followed by moderate-severe white matter hyperintensities (45), lacunes (41), and microbleeds (30). Recurrent stroke rates did not strictly increase with increasing SVD score, i.e. rates were 2.4%/pt-yr (95% CI 1.5, 3.9), 1.4 (0.8, 2.3), 2.0 (1.3, 3.2), 3.8 (2.5, 5.7), and 3.2 (1.8, 5.6) respectively. When SVD scores of 0-2 vs. 3-4 were grouped and compared with two other models for predicting recurrent stroke in this cohort, the SVD score model did not outperform. (Table) Conclusions: SVD score features were very common in the SPS3 cohort. Higher (3-4) vs. lower (0-2) SVD scores predicted recurrent stroke with similar predictive ability to models including clinical risk factors only +/- fewer MRI features. Further testing of the SVD score is warranted.

Prediction of recurrent stroke with ABCD2 and ABCD3 scores in patients with symptomatic 50-99% carotid stenosis.

BackgroundAlthough it is preferable that all patients with a recent Transient Ischemic Attack (TIA) undergo acute carotid imaging, there are centers with limited access to such acute examinations. It is controversial whether ABCD2 or ABCD3 scores can be used to triage patients to acute or delayed carotid imaging. It would be acceptable that some patients with a symptomatic carotid stenosis are detected with a slight delay as long as those who will suffer an early recurrent stroke are detected within 24 hours. The aim of this study is to analyze the ability of ABCD2 and ABCD3 scores to predict ipsilateral ischemic stroke among patients with symptomatic 50-99% carotid stenosis.MethodsIn this secondary analysis of the ANSYSCAP-study, we included 230 consecutive patients with symptomatic 50-99% carotid stenosis. We analyzed the risk of recurrent ipsilateral ischemic stroke before carotid endarterectomy based on each parameter of the ABCD2 and ABCD3 scores separately, and for total ABCD2 and ABCD3 scores. We used Kaplan-Meier analysis.ResultsNone of the parameters in the ABCD2 or ABCD3 scores could alone predict all 12 of the ipsilateral ischemic strokes that occurred within 2 days of the presenting event, but clinical presentation tended to be a statistically significant risk factor for recurrent ipsilateral ischemic stroke (p = 0.06, log rank test). An ABCD2 score ≥2 and an ABCD3 score ≥4 could predict all 12 of these strokes as well as all 25 ipsilateral ischemic strokes that occurred within 14 days. To use ABCD3 score seems preferable over the ABCD2 score because a higher proportion of low risk patients were identified (17% of the patients had an ABCD3 score <4 while only 6% had an ABCD2 < 2).ConclusionsAlthough it is preferable that carotid imaging be performed within 24 hours, our data support that an ABCD3 score ≥4 might be used for triaging patients to acute carotid imaging in clinical settings with limited access to carotid imaging. However, our findings should be validated in a larger cohort study.

Prediction of recurrent stroke and myocardial infarction after stroke: a systematic review of clinical prediction models

We identified twelve model development studies and eleven evaluation studies. Investigators often did not report effective sample size, regression coefficients, handling of missing data; typically categorised continuous predictors; and used data dependent methods to build models (e.g., univariate screening of predictors). Four models were evaluated. The pooled area under the receiver operating characteristic curve (AUROCC) estimate for the Essen Stroke Risk Score (ESRS) was 0.60 (95% CI 0.59 to 0.62, ten studies), for the Stroke Prognosis Instrument II (SPI-II) was 0.62 (95% CI 0.60 to 0.64, nine studies) and a single study of the Recurrence Risk Estimator at 90 days (RRE-90) was 0.72 (95% CI 0.56 to 0.88, one study) and of the Life Long After Cerebral ischemia (LiLAC) was 0.65 (95% CI 0.61 to 0.69, one study). Conclusions The available models for recurrent stroke discriminate only modestly between patients with and without a recurrent stroke or MI. Performance may be improved by addressing commonly encountered methodological flaws.

Vascular Risk Prediction in Ischemic Stroke Patients Undergoing in-patient Rehabilitation – Insights from the Investigation of Patients with Ischemic Stroke in Neurologic Rehabilitation (INSIGHT) Registry

In-patient rehabilitation following ischemic stroke offers a unique opportunity for risk factor and lifestyle modification. Quantification of risk in this setting may help to tailor therapy, increase physician awareness and patient compliance and thus to reduce recurrent vascular events. To validate the predictive value of established secondary stroke risk scores. One thousand one hundred sixty-three patients undergoing in-patient rehabilitation following recent ischemic stroke in 15 German neurologic rehabilitation centers were included 0·9 ± 0·5 months after the index event. Outcome information was available for 846 participants (72·7%) after a mean follow-up of 13 ± 2·3 months. Patients' mean age was 66·3 ± 12·3 years and 42·5% were women. The National Institutes of Health Stroke Scale (mean 4·0 ± 3·9), modified Rankin scale (median 2, range 0-5), and Barthel Index (median 90, range 0-100) indicated good functional status. A recurrent fatal or nonfatal stroke during follow-up occurred in 6·7% and combined vascular events (stroke, myocardial infarction, vascular death) in 10·9%. The predictive accuracy for recurrent stroke was slightly higher on the Essen Stroke Risk Score than on the Stroke Prognostis Instrument II (area under the curve 0·62 vs. 0·56), while both scores had a similar predictive accuracy for combined vascular events. Risk stratification on the Essen Stroke Risk Score and Stroke Prognostis Instrument II provides a moderate accuracy for the prediction of recurrent stroke and vascular events in patients undergoing neurologic in-patient rehabilitation. Although individual risk prediction may remain imprecise, the use of these scores should be encouraged.

Abstract 70: Predicting Recurrent Stroke After TIA And Minor Stroke: Results Of The Prospective CT And MRI In The Triage Of TIA And Minor Cerebrovascular Events To Identify High Risk Patients (CATCH) Study.

Background. TIA and minor stroke have a substantial risk of recurrent stroke, particularly within the first 48 hours. There is therefore a need to identify the highest risk patients urgently to implement early treatments. Imaging using MRI can identify patients at high risk for a recurrent stroke. However MRI is not available emergently in many institutions. If CT/CTA could identify high-risk patients then this would be more widely applicable. Methods. 510 consecutive TIA and minor stroke patients who were assessed by a stroke neurologist and had a CT/CTA completed within 24 hours of onset were prospectively enrolled. 420 had baseline brain MRI completed also. We used multiple imputation based on a previous meta-analysis of predictors of DWI positivity (motor or speech symptoms, atrial fibrillation, symptomatic carotid stenosis &gt;=50% and symptom duration greater than 60 minutes) to impute MRI results for patients without baseline MRI. We assessed the risk of recurrent stroke within 90 days using standard clinical variables and predefined abnormalities on the CT/CTA at risk metric (acute ischemia on CT and/or intracranial or extracranial occlusion or stenosis &gt;=50%) and MRI (DWI positivity). Results. There were 36 recurrent strokes (7.1% 95%CI: 5.0-9.6). Median time to event was 1 day (IQR 7.5). Median time from symptom onset to CTA was 5.5 hours (IQR: 6.4 hours), median time to MRI was 17.5 hours (IQR: 12 hours). In the univariate analysis; symptoms ongoing at first assessment, HR 2.2 (95%CI: 1.02-4.9), CT/CTA at risk metric, HR 4·0 (95%CI: 2·0-8·5) and DWI positivity 3.2 (1.3-7.6) predicted recurrent stroke. In the multivariable analysis only CT/CTA at risk metric (OR 3.6 (1.7-7.5, p=0.001) and DWI positivity (OR 2.3 (0.9-5.8) predicted recurrent stroke. Diagnostic accuracy of CT/CTA in predicting recurrent stroke was: sensitivity 67%, specificity 68%, PPV 14%, NPV 96%. Diagnostic accuracy of MRI: sensitivity 83%, specificity 40%, PPV 10%, NPV 97%. Using ROC analysis CT/CTA and MRI were not significantly different in their accuracy in prediction of recurrent stroke (0.67 versus 0.61, p=0.18). Conclusions. Early assessment of the intracranial and extracranial vasculature using CT/CTA predicts recurrent stroke and clinical outcome in patients with TIA and minor stroke. There is a trade off between sensitivity and specificity when CTA and MRI are compared. In many institutions CTA is more quickly available than MRI and given a median time to event of one day, physicians should access whichever technique is available quicker in their institution to allow implementation of aggressive secondary prevention treatments in appropriate patients.

Prediction of Recurrent Stroke and Vascular Death in Patients With Transient Ischemic Attack or Nondisabling Stroke

Several predictive scores have been developed and validated for stratifying cerebrovascular patients based on the risk of future (cerebro)vascular events. We aimed to prospectively compare the predictive accuracy of the Essen Stroke Risk Score, Stroke Prognostic Instrument, Hankey score, and the Life Long After Cerebral ischemia score. Between August 2005 and December 2006, we included 2381 patients from 10 German stroke centers with an acute nondisabling ischemic stroke or transient ischemic attack and with prospective assessment of clinical variables for calculation of the predictive scores. A total of 1897 patients (79.7%) could be followed up for a median of 1 year. To evaluate the performance of each model, we calculated the area under the curve by receiver operating characteristic. In addition, we used the recommended cutoff values for calculation of sensitivity and specificity for stroke or the combined outcome of stroke or cardiovascular death. The Kaplan-Meier estimate for the overall annual stroke risk was 4.8% and for recurrent stroke or cardiovascular death 6.6%. We could confirm the predictive value of all 4 previously developed scores with a marginally superior performance of the SPI-II. In patients with acute nondisabling ischemic stroke or transient ischemic attack, all 4 scores are able to stratify the risk of recurrent stroke or the combined outcome. Simple point scores (Essen Stroke Risk Score, Stroke Prognostic Instrument) may help to raise awareness for medical prevention in clinical routine and increase compliance with risk factor modification.

Inflammatory Markers and Poor Outcome after Stroke: A Prospective Cohort Study and Systematic Review of Interleukin-6

The objective of this study was to determine whether: (a) markers of acute inflammation (white cell count, glucose, interleukin-6, C-reactive protein, and fibrinogen) are associated with poor outcome after stroke and (b) the addition of markers to previously validated prognostic models improves prediction of poor outcome. We prospectively recruited patients between 2002 and 2005. Clinicians assessed patients and drew blood for inflammatory markers. Patients were followed up by postal questionnaire for poor outcome (a score of>2 on the modified Rankin Scale) and death through the General Register Office (Scotland) at 6 mo. We performed a systematic review of the literature and meta-analysis of the association between interleukin-6 and poor outcome after stroke to place our study in the context of previous research. We recruited 844 patients; mortality data were available in 844 (100%) and functional outcome in 750 (89%). After appropriate adjustment, the odds ratios for the association of markers and poor outcome (comparing the upper and the lower third) were interleukin-6, 3.1 (95% CI: 1.9-5.0); C-reactive protein, 1.9 (95% CI: 1.2-3.1); fibrinogen, 1.5 (95% CI: 1.0-2.36); white cell count, 2.1 (95% CI: 1.3-3.4); and glucose 1.3 (95% CI: 0.8-2.1). The results for interleukin-6 were similar to other studies. However, the addition of inflammatory marker levels to validated prognostic models did not materially improve model discrimination, calibration, or reclassification for prediction of poor outcome after stroke. Raised levels of markers of the acute inflammatory response after stroke are associated with poor outcomes. However, the addition of these markers to a previously validated stroke prognostic model did not improve the prediction of poor outcome. Whether inflammatory markers are useful in prediction of recurrent stroke or other vascular events is a separate question, which requires further study. Please see later in the article for the Editors' Summary.

Risk stratification and prediction of recurrent stroke

Approximately 30% of patients suffer a recurrent stroke during the 5 years after their first cerebrovascular event. Predictors differ for short- and long-term risk of stroke recurrence. Atherosclerotic disease is the single most important predictor for short-term recurrence, while most classical cardiovascular risk factors, as well as age, are independent predictors for long-term recurrence. Several prognostic models for long-term risk of recurrent stroke have been validated, but the predictive value and applicability of these models remains limited. However, an improved understanding of the risk for recurrent stroke is vital for the design of secondary prevention trials as well as individual treatment decisions, especially if the medication or procedure is expensive or risky. The aim of this article is to elucidate the relationship between stroke risk factors, prognostic scores and their predictive value in recurrent stroke.