In-hospital dynamics of glucose, blood pressure and temperature predict outcome in patients with acute ischaemic stroke

Machine learning-based prognostication of mortality in stroke patients

Objectives

Predicting stroke mortality is crucial for personalized care. This study aims to design and evaluate a machine learning model to predict one-year mortality after a stroke.

Materials and methods

Data from the National Multiethnic Stroke Registry was utilized. Eight machine learning (ML) models were trained and evaluated using various metrics. SHapley Additive exPlanations (SHAP) analysis was used to identify the influential predictors.

Results

The final analysis included 9840 patients diagnosed with stroke were included in the study. The XGBoost algorithm exhibited optimal performance with high accuracy (94.5%) and AUC (87.3%). Core predictors encompassed National Institutes of Health Stroke Scale (NIHSS) at admission, age, hospital length of stay, mode of arrival, heart rate, and blood pressure. Increased NIHSS, age, and longer stay correlated with higher mortality. Ambulance arrival and lower diastolic blood pressure and lower body mass index predicted poorer outcomes.

Conclusions

This model's predictive capacity emphasizes the significance of NIHSS, age, hospital stay, arrival mode, heart rate, blood pressure, and BMI in stroke mortality prediction. Specific findings suggest avenues for data quality enhancement, registry expansion, and real-world validation. The study underscores machine learning's potential for early mortality prediction, improving risk assessment, and personalized care. The potential transformation of care delivery through robust ML predictive tools for Stroke outcomes could revolutionize patient care, allowing for personalized plans and improved preventive strategies for stroke patients. However, it is imperative to conduct prospective validation to evaluate its practical clinical effectiveness and ensure its successful adoption across various healthcare environments.

Predicting 90-Day Prognosis in Ischemic Stroke Patients Post Thrombolysis Using Machine Learning

(1) Objective: This study aimed to construct a machine learning model for predicting the prognosis of ischemic stroke patients who underwent thrombolysis, assessed through the modified Rankin Scale (mRS) score 90 days after discharge. (2) Methods: Data were sourced from Qatar's stroke registry covering January 2014 to June 2022. A total of 723 patients with ischemic stroke who had received thrombolysis were included. Clinical variables were examined, encompassing demographics, stroke severity indices, comorbidities, laboratory results, admission vital signs, and hospital-acquired complications. The predictive capabilities of five distinct machine learning models were rigorously evaluated using a comprehensive set of metrics. The SHAP analysis was deployed to uncover the most influential predictors. (3) Results: The Support Vector Machine (SVM) model emerged as the standout performer, achieving an area under the curve (AUC) of 0.72. Key determinants of patient outcomes included stroke severity at admission; admission systolic and diastolic blood pressure; baseline comorbidities, notably hypertension (HTN) and coronary artery disease (CAD); stroke subtype, particularly strokes of undetermined origin (SUO); and hospital-acquired urinary tract infections (UTIs). (4) Conclusions: Machine learning can improve early prognosis prediction in ischemic stroke, especially after thrombolysis. The SVM model is a promising tool for empowering clinicians to create individualized treatment plans. Despite limitations, this study contributes to our knowledge and encourages future research to integrate more comprehensive data. Ultimately, it offers a pathway to improve personalized stroke care and enhance the quality of life for stroke survivors.

Efficacy and Safety of Analog Insulin in Comparison With Human Insulin for Hyperglycemia in Hospitalized Patients With Acute Stroke: A Randomized, Open-Label, Single-Center Trial

OBJECTIVE

To assess the efficacy and safety of analog insulins in comparison with human insulins for hyperglycemia in hospitalized patients with acute stroke.

METHODS

In this single-center, open-label, randomized trial, 102 patients (age 59.4 ± 11.7 years, 54 women) admitted with acute stroke (52 ischemic, 50 hemorrhagic) and hyperglycemia were assigned to analog insulin (n = 52) or human insulin (n = 50) group during February to June 2021. Insulin was initiated and titrated according to the predefined standard protocol. The capillary blood glucose (BG) level was monitored by standardized glucometers. The primary outcomes were mean daily BG and the number of hypoglycemic events.

RESULTS

Between the 2 treatment groups, there was no significant difference in the mean daily BG (P >.05 for all days) or in the frequency of hypoglycemic episodes (P =.727). Four participants experienced severe hypoglycemia; all were receiving human insulin (P =.054). In the analog insulin group, there was a tendency toward lower daily total requirement for insulin (P =.053). The difference in bolus insulin dose was significantly lower in the analog insulin group (P =.029), but the difference in basal insulin dose was similar (P =.167). Between the 2 groups, there were no significant differences in the hospital mortality rate, modified Rankin Scale score on outcome, or length of hospital stay (P =.729,.658, and.918, respectively).

CONCLUSION

Hospitalized patients acute stroke and hyperglycemia exhibited similar mean BG but a trend of lower incidence of severe hypoglycemia when treated with analog insulins in comparison with human insulin.

The neurovascular unit and systemic biology in stroke - implications for translation and treatment

Ischaemic stroke is a leading cause of disability and death for which no acute treatments exist beyond recanalization. The development of novel therapies has been repeatedly hindered by translational failures that have changed the way we think about tissue damage after stroke. What was initially a neuron-centric view has been replaced with the concept of the neurovascular unit (NVU), which encompasses neuronal, glial and vascular compartments, and the biphasic nature of neural-glial-vascular signalling. However, it is now clear that the brain is not the private niche it was traditionally thought to be and that the NVU interacts bidirectionally with systemic biology, such as systemic metabolism, the peripheral immune system and the gut microbiota. Furthermore, these interactions are profoundly modified by internal and external factors, such as ageing, temperature and day-night cycles. In this Review, we propose an extension of the concept of the NVU to include its dynamic interactions with systemic biology. We anticipate that this integrated view will lead to the identification of novel mechanisms of stroke pathophysiology, potentially explain previous translational failures, and improve stroke care by identifying new biomarkers of and treatment targets in stroke.

Analysis of severity in ischemic stroke patients with coronary artery disease in the telestroke network

Aim: This study investigated demographic and clinical risk factors associated with severity in acute ischemic stroke (AIS) patients with a history of coronary artery disease (CAD). Materials & methods: Multivariate logistic regression models were used to determine specific factors associated with worsening or improving neurologic functions among AIS patients with CAD treated in the telestroke. Results: AIS patients with CAD, peripheral vascular disease (OR: 3.995; 95% CI: 1.035-15.413; p = 0.044) and hypertension (OR: 6.056; 95% CI: 1.207-30.374; p = 0.029) were associated with worsening neurologic functions. Conclusion: Our findings suggest the need to develop strategic management to improve the care of AIS patients with a history of CAD that present with hypertension and peripheral vascular disease in the telestroke network.

24-hour blood pressure variability and treatment effect of intravenous alteplase in acute ischaemic stroke

Introduction

To assess the association between 24 h blood pressure variability (BPV) on functional outcome and treatment effect of intravenous alteplase in acute ischaemic stroke.

Patients and methods

In all patients with acute ischaemic stroke of unknown onset randomised in the WAKE-UP (Efficacy and Safety of magnetic resonance imaging [MRI]-based Thrombolysis in Wake-Up Stroke) trial, blood pressure (BP) was measured before randomisation and after initiation of treatment at regular intervals up to 24 hours. Individual BPV was measured by coefficient of variation (CV) of all BP values. Primary outcome measure was favourable outcome defined by a modified Rankin Scale (mRS) score 0 or 1 at 90 days after stroke.

Results

BP measurements were available for 498 of 503 patients randomised (177 women [35.5%], mean age [SD] of 65.2 [11.5] years). Systolic BPV was not associated with the treatment effect of thrombolysis (test for interaction, p = 0.46). The adjusted odds ratio (aOR) for favourable outcome with alteplase, adjusted for age, stroke severity and baseline BP on admission, did not show an association across the quintiles of increasing systolic BPV with an aOR 1.89 (95% confidence interval [CI], 0.76-4.70) in the lowest quintile to aOR 1.05 (95% CI, 0.43-2.56) in the highest quintile. Higher mean systolic BP was associated with a smaller treatment effect of thrombolysis with a significant interaction (p = 0.033). The aOR for favourable outcome with alteplase decreased with quintiles of increasing mean systolic BP from aOR 3.16 (95% CI, 1.26-7.93) in the lowest quintile to aOR 0.84 (95% CI, 0.34-2.10) in in the highest quintile.

Conclusions

There was a significant interaction between mean systolic BP and treatment effect of thrombolysis with higher mean systolic BP being associated with poorer outcome. BPV was not associated with outcome after thrombolysis.ClinicalTrials.gov identifier NCT01525290.

Verapamil as an Adjunct Therapy to Reduce tPA Toxicity in Hyperglycemic Stroke: Implication of TXNIP/NLRP3 Inflammasome

Thrombolytic therapy has remained quite challenging in hyperglycemic patients for its association with poor prognosis and increased hemorrhagic conversions. We recently showed that tissue plasminogen activator (tPA)-induced cerebrovascular damage is associated with thioredoxin-interacting protein (TXNIP) upregulation, which has an established role in the detrimental effects of hyperglycemia. In the present work, we investigated whether verapamil, an established TXNIP inhibitor, may provide protection against hyperglycemic stroke and tPA-induced blood-brain barrier (BBB) disruption. Acute hyperglycemia was induced by intraperitoneal administration of 20% glucose, 15 min prior to transient middle cerebral artery occlusion (tMCAO). Verapamil (0.15 mg/kg) or saline was intravenously infused with tPA at hyperglycemic reperfusion, 1 h post tMCAO. After 24 h of ischemia/reperfusion (I/R), mice were assessed for neurobehavioral deficits followed by sacrifice and evaluation of brain infarct volume, edema, and microbleeding. Alterations in TXNIP, inflammatory mediators, and BBB markers were further analyzed using immunoblotting or immunostaining techniques. As adjunctive therapy, verapamil significantly reduced tPA-induced BBB leakage, matrix metalloproteinase 9 (MMP-9) upregulation, and tight junction protein deregulation, which resulted in lesser hemorrhagic conversions. Importantly, verapamil strongly reversed tPA-induced TXNIP/NLRP3 (NOD-like receptor pyrin domain-containing-3) inflammasome activation and reduced infarct volume. This concurred with a remarkable decrease in high-mobility group box protein 1 (HMGB-1) and nuclear factor kappa B (NF-κB) stimulation, leading to less priming of NLRP3 inflammasome. This preclinical study supports verapamil as a safe adjuvant that may complement thrombolytic therapy by inhibiting TXNIP's detrimental role in hyperglycemic stroke.

Blood Pressure and Hospital Discharge Outcomes in Acute Ischemic Stroke Patients Undergoing Reperfusion Therapy

INTRODUCTION

Acute management of blood pressure in ischemic stroke treated with reperfusion therapy remains uncertain. We evaluated blood pressures during the first 24-hours after reperfusion therapy in relation to in-hospital outcomes.

METHODS

We conducted a single-center retrospective study of blood pressure in the first 24 hours among ischemic stroke patients who underwent reperfusion therapy with intravenous thrombolysis (IVT) or mechanical thrombectomy (MT) at a tertiary referral center. Blood pressure variability was expressed as the range between the highest and the lowest pressures. Outcomes of interest were discharge disposition and in-hospital mortality. Favorable outcome was defined as a discharge destination to home or inpatient rehabilitation facility (IRF). Multivariable logistic regression analysis was performed with adjustment for age, National Institutes of Health Stroke Scale score, and patients receiving reperfusion therapy.

RESULTS

Among the 140 ischemic stroke patients (117 IVT, 84 MT and 61 both), 95 (67.8%) had favorable discharge disposition and 24 (17.1%) died. Higher 24-hour peak systolic blood pressures (SBPs) and peak mean arterial pressures (MAPs) were independently associated with a lower likelihood of favorable discharge disposition, with an adjusted odds ratio (aOR) 0.868, 95 % CI 0.760 - 0.990 per 10 mm Hg for SBP and aOR 0.710, 95% CI 0.515 - 0.980 for MAP, and with increased odds of death aOR 1.244, 95% CI 1.056-1.467 and aOR 1.760, 95% CI 1.119 - 2.769 respectively. Greater variability of SBP and MAP was also associated with odds of death aOR 1.327, 95% CI 1.104 - 1.595 and aOR 1.577, 95% CI 1.060- 2.345 respectively, without a significant effect on discharge disposition.

CONCLUSION

In the first 24 hours after reperfusion therapy, higher peak and variable blood pressures are associated with unfavorable discharge outcomes and increased in-hospital mortality. Further studies in stroke patients undergoing reperfusion therapy might target blood pressure reduction and variability to improve patient outcomes.