Early identification of potentially salvageable tissue with MRI-based predictive algorithms after experimental ischemic stroke

Individualized stroke treatment decisions can be improved by accurate identification of the extent of salvageable tissue. Magnetic resonance imaging (MRI)-based approaches, including measurement of a ‘perfusion-diffusion mismatch'' and calculation of infarction probability, allow assessment of tissue-at-risk; however, the ability to explicitly depict potentially salvageable tissue remains uncertain. In this study, five predictive algorithms (generalized linear model (GLM), generalized additive model, support vector machine, adaptive boosting, and random forest) were tested in their potency to depict acute cerebral ischemic tissue that can recover after reperfusion. Acute T₂-, diffusion-, and perfusion-weighted MRI, and follow-up T₂ maps were collected from rats subjected to right-sided middle cerebral artery occlusion without subsequent reperfusion, for training of algorithms (Group I), and with spontaneous (Group II) or thrombolysis-induced reperfusion (Group III), to determine infarction probability-based viability thresholds and prediction accuracies. The infarction probability difference between irreversible—i.e., infarcted after reperfusion—and salvageable tissue injury—i.e., noninfarcted after reperfusion—was largest for GLM (20±7%) with highest accuracy of risk-based identification of acutely ischemic tissue that could recover on subsequent reperfusion (Dice''s similarity index=0.79±0.14). Our study shows that assessment of the heterogeneity of infarction probability with MRI-based algorithms enables estimation of the extent of potentially salvageable tissue after acute ischemic stroke.     

MRI-based age prediction using hidden Markov models

Cortical thinning and intracortical gray matter volume losses are widely observed in normal ageing, while the decreasing rate of the volume loss in subjects with neurodegenerative disorders such as Alzheimer's disease is reported to be faster than the average speed. Therefore, neurodegenerative disease is considered as accelerated ageing. Accurate detection of accelerated ageing based on the magnetic resonance imaging (MRI) of the brain is a relatively new direction of research in computational neuroscience as it has the potential to offer positive clinical outcome through early intervention. In order to capture the faster structural alterations in the brain with ageing, we propose in this paper a computational approach for modelling the MRI-based structure of the brain using the framework of hidden Markov models, which can be utilized for age prediction. Experiments were carried out on healthy subjects to validate its accuracy and its robustness. The results have shown its ability of predicting the brain age with an average normalized age-gap error of two to three years, which is superior to several recently developed methods for brain age prediction.     

Aphasia assessment and functional outcome prediction in patients with aphasia after stroke

Available studies did not clarify whether a language examination may predict functional and motor outcome in patients with aphasia undergoing rehabilitation. This was the aim of the current study. Language examination considered in this study was the Aachen Aphasia Test (AAT). One hundred fifty-six patients with a primary diagnosis of acute cerebrovascular accident of left hemisphere were included: 105 with and 51 without aphasia. Backward stepwise regression analysis was used to predict final scores in total-, motor-, and cognitive-Functional Independence Measure (FIM). The independent variables were age, gender, stroke type, stroke lesion size, onset to admission interval, National Institute of Health Stroke Scale, Fugl-Meyer Scale, Trunk Control Test, initial motor-FIM, and AAT (spontaneous speech, token test, repetition, written language, confrontation naming, comprehension). In the multivariate regression analysis, comprehension was the only function of the language to be predictor of the final total-FIM (β = +0.35) and final cognitive-FIM (β = +0.61). Comprehension was a predictor of total-FIM (β = +0.27) and cognitive-FIM (β = +0.54) as well, when additional backward stepwise regression analysis (which did not include comprehension and expression scores in the final total- and cognitive-FIM) was performed. When multivariate regression analysis took into consideration only language functions as independent variables, spontaneous speech (β = +0.41) only was the predictor of motor-FIM. The study highlights that AAT is the predictor of functional outcome in the patient with aphasia. Among the functions of language, comprehension seems to be the most important predictive factor of total- and cognitive-FIM, while spontaneous speech seems to be a predictor of motor-FIM.     

Brief cognitive assessment and prediction of functional outcome in stroke

To evaluate the ability to predict outcome with a brief measure of cognitive ability, we tested consecutive admissions who received inpatient rehabilitation for stroke with the Repeatable Battery for Assessment of Neuropsychological Symptoms (RBANS). Six months later, 34 discharged patients were contacted by telephone and were interviewed using a battery of functional outcome and quality of life measures. Multiple regression analysis showed that inpatient RBANS indexes predicted cognitive disability 6 months later. The present findings support the use of cognitive evaluations of patients with acute stroke to assist with prediction of outcome to be used in treatment planning.     

MRI-based intravenous thrombolysis in stroke patients with unknown time of symptom onset

Background: Currently, stroke patients with unknown time of symptom onset (UTOS) are excluded from therapy with intravenous tissue Plasminogen Activator. We hypothesized that MRI‐based intravenous thrombolysis is safe in UTOS. Methods: We analyzed radiological and clinical data as well as outcomes of stroke patients (including UTOS) who received intravenous thrombolytic therapy after MRI. Results: Compared to patients with known time of symptom onset (n = 131), UTOS (n = 17) were older (81, 71–88 vs. 75 years, 66–82, P = 0.03), had a longer median time between last‐seen‐well and thrombolysis (12.3 h, IQR 11.5–15.2 h vs. 2.1 h, 1.8–2.8 h, P < 0.01), had a longer median door‐to‐needle time (86 min, 49–112 vs. 60 min, 49–76, P = 0.02), and a higher rate of arterial obstruction on MR‐angiography (82.4% vs. 56.5%, P = 0.04). No symptomatic intracerebral hemorrhage occurred in UTOS. After 3 months, there was no significant difference between groups concerning good functional outcome (modified Rankin Scale 0–2; 35.3% vs. 49.6%, P = 0.26) or mortality (0% vs. 15.3%, P = 0.08). In multivariate analyses including age, gender, baseline NIHSS, and atrial fibrillation UTOS did not have an independent effect on good functional outcome after 3 months (OR 1.16; 0.32–4.12, P = 0.81). Conclusions: Thrombolysis after MRI seems safe and effective in UTOS. This observation may encourage those who plan prospective placebo‐controlled trials of thrombolytics in this subgroup of stroke patients.     

Hypertension and hyperglycemia in experimental stroke.

Both diabetes mellitus and hypertension are risk factors for stroke and also influence prognosis following stroke. Experimentally, hyperglycemia augments cortical infarct size in stroke models where collateral circulation exists, and infarct size in hypertensive rats is larger than in normotensive strains. Whether the deleterious effect of hyperglycemia is altered in the setting of hypertension has not previously been studied experimentally. The effect of hyperglycemia on infarct size in spontaneously hypertensive rats was examined in this study. Focal neocortical cerebral ischemia was induced by tandem right common carotid and middle cerebral artery occlusion. Preischemic hyperglycemia had no influence on infarct volume whether the duration of postischemic hyperglycemia was transient or prolonged. Although hyperglycemia increases infarct size in cortical stroke models where collateral circulation is available, this study demonstrates the effect can be modified by the presence of underlying hypertension.     

随机森林和决策树模型在轻型缺血性脑卒中患者复发预测中的应用分析

目的 构建轻型缺血性脑卒中（MIS）患者2年内复发的随机森林和决策树预测模型,并分析模型的临床应用价值。方法 回顾性收集2020年7月1日至12月31日于山西省心血管病医院神经内科就诊的520例MIS患者的病历资料,根据2年内是否复发将患者分为复发组和未复发组。基于缺失森林对数据进行填补,根据文献检索与专家讨论结果筛选预测变量并进行单因素分析,合成少数过采样技术-标称连续（SMOTE-NC）技术处理数据不平衡,采用贝叶斯优化十折交叉验证构建随机森林、决策树模型并与Logistic回归模型进行比较。基于受试者工作特征曲线下面积（AUC）、布里尔分数（BS）与校准曲线分别评价模型的区分度与校准度。对预测性能最好的模型采用SHAP模型解释预测结果。结果 2年内复发患者共93例（17.9%）。两组患者的年龄,吸烟、糖尿病、循环梗死部位、多发性脑梗死比例,以及舒张压、红细胞压积、血小板计数、低密度脂蛋白水平比较,差异有统计学意义（P<0.05）。Logistic回归模型、决策树模型与随机森林模型在测试集中,预测MIS患者2年内复发情况的AUC(95%CI)分别为0.764(0.691～0....     

Infarct volume, neurological severity and PAF binding to platelets of patients with acute cerebral ischemic stroke.

Studies show that Platelet Activating Factor (PAF) is involved in the cerebrovascular response to ischemia, and that its binding to platelets may change in stroke victims. The purpose of this study was to determine whether binding of PAF to platelets of stroke patients could serve as an index for determining the volume of ischemic strokes and severity of neurological presentation. Thirteen stroke patients and 21 healthy controls were studied. The neurological severity of these stroke patients was evaluated by the Scandinavian Stroke Scale. Infarct volume was assessed by planimetric measures of brain CT. PAF binding to platelets was determined by use of radiolabelled PAF. (3H)PAF binding to platelets of stroke patients was lower than in controls (149.58 +/- 46.11 and 212.1 +/- 10.3 receptors cell-1, respectively, p < 0.001) and was significantly correlated with infarct volume (r = -0.606, p = 0.014) and with neurological score (r = 0.527, p = 0.032). No correlation was observed between neurological score and infarct volume. The study confirms the involvement of PAF in the pathogenesis of brain ischemia and neuronal damage. It shows that PAF binding to platelets of stroke patients correlates both with the extent of neuronal damage and the associated neurological impairment, and may serve as an additional index in the assessment of stroke severity and clinical outcome of stroke victims.     

Comparison of different intravascular thread occlusion models for experimental stroke in rats

For the induction of ischemic strokes of varying sizes in rats, different types of threads were used to occlude the middle cerebral artery (MCA) in combination with or without the posterior communicating artery (PCOM) and the common carotid artery (CCA). During vessel occlusion brain tissue partial oxygen pressure (ptiO2) and regional cerebral blood flow (rCBF) were monitored using a combined ptiO2/laser Doppler flow probe. Following neurological assessment animals were sacrificed at 3, 8 and 24 h and the necrotic volume was measured on serial coronary slices. In another experimental group, rCBF was measured 1h post-insult with the iodo[14C]antipyrine method. Animals with selective MCA occlusion showed less reduction of ptiO2 and rCBF and smaller infarcts when compared with animals with combined occlusion of the MCA, CCA and PCOM. Both groups, selective MCA occlusion and combined occlusion of the MCA, CCA and PCOM, demonstrated a high reproducibility and low variability of stroke size. Relative growth of stroke size within 24 h was higher in animals with selective MCA occlusion. Therefore, the selective MCAO model may be advantageous for studies of neuroprotective strategies.     

Long-term behavioral assessment of function in an experimental model for ischemic stroke

Middle cerebral artery occlusion (MCAO) in rats is a well-studied experimental model for ischemic stroke leading to brain infarction and functional deficits. Many preclinical studies have focused on a small time window after the ischemic episode to evaluate functional outcome for screening therapeutic candidates. Short evaluation periods following injury have led to significant setbacks due to lack of information on the delayed effects of treatments, as well as short-lived and reversible neuroprotection, so called false-positive results. In this report, we evaluated long-term functional deficit for 90 days after MCAO in two rat strains with two durations of ischemic insult, in order to identify the best experimental paradigm to assess injury and subsequent recovery. Behavioral outcomes were measured pre-MCAO followed by weekly assessment post-stroke. Behavioral tests included the 18-point composite neurological score, 28-point neuroscore, rearing test, vibrissae-evoked forelimb placing test, foot fault test and the CatWalk. Brain lesions were assessed to correlate injury to behavior outcomes at the end of study. Our results indicate that infarction volume in Sprague-Dawley rats was dependent on occlusion duration. In contrast, the infarction volume in Wistar rats did not correlate with the duration of ischemic episode. Functional outcomes were not dependent on occlusion time in either strain; however, measurable deficits were detectable long-term in limb asymmetry, 18- and 28-point neuroscores, forelimb placing, paw swing speed, and gait coordination. In conclusion, these behavioral assays, in combination with an extended long-term assessment period, can be used for evaluating therapeutic candidates in preclinical models of ischemic stroke.     

Chronic metformin treatment improves post‐stroke angiogenesis and recovery after experimental stroke

Metformin is currently the first‐line treatment drug for type 2 diabetes. Metformin is a well‐known activator of AMP‐activated protein kinase (AMPK). In experimental studies, metformin has been shown to exert direct vascular effects by increasing vascular endothelial growth factor expression and improving microvascular density. As stroke is the leading cause of long‐term disability and angiogenesis is implicated as an important mechanism in functional recovery, we hypothesized that chronic metformin treatment would improve post‐stroke functional recovery by enhancing functional microvascular density. For this study, C57BL/6N male mice were subjected to a 60‐min middle cerebral artery occlusion, and were given 50 mg/kg/day metformin beginning 24 h post‐stroke for 3 weeks. Behavioral recovery was assessed using adhesive‐tape removal and the apomorphine‐induced turning test. The role of angiogenesis was assessed by counting vessel branch points from fluorescein‐conjugated lectin‐perfused brain sections. Importantly even if metformin treatment was initiated 24 h after injury it enhanced recovery and significantly improved stroke‐induced behavioral deficits. This recovery occurred in parallel with enhanced angiogenesis and with restoration of endogenous cerebral dopaminergic tone and revascularization of ischemic tissue. We assessed if the effects on recovery and angiogenesis were mediated by AMPK. When tested in AMPK α‐2 knockout mice, we found that metformin treatment did not have the same beneficial effects on recovery and angiogenesis, suggesting that metformin‐induced angiogenic effects are mediated by AMPK. The results from this study suggest that metformin mediates post‐stroke recovery by enhancing angiogenesis, and these effects are mediated by AMPK signaling.     

Streptokinase treatment versus calcium overload blockade in experimental thromboembolic stroke

Thromboembolic brain ischemia was produced in dogs using an autologous blood clot model. The effect of postembolic treatment with flunarizine and streptokinase on hemispheric cerebral metabolic rate for oxygen (CMRO2), oxygen extraction ratio (OER), and cerebral blood flow (CBF) was studied by positron emission tomography (oxygen-15 technique) 24 hours after the insult. We studied five groups of experimental dogs and compared them with a control group of nonembolized dogs. Group I received no treatment, Group II was treated locally with 500,000 IU streptokinase starting 30 minutes after the insult, Group III received streptokinase locally 30 minutes after the insult and 0.1 mg/kg i.v. flunarizine immediately after the insult and 2 hours later, Group IV received flunarizine as Group III, and Group V was orally pretreated with 0.5 mg/kg/day flunarizine during 2 weeks preceding embolization. Compared with the contralateral hemisphere, in the embolized hemisphere a significant reduction of CMRO2 (-25% to -40%) and CBF in normocapnia (-35%) and hypercapnia (-50%) was observed in Groups I, II, and V. In Groups III and IV, CMRO2, OER, and CBF of the embolized hemisphere were within the normal range during normocapnia and hypercapnia; the extent of the ischemic lesions was markedly less than in the other groups of experimental dogs. We conclude that flunarizine treatment after experimental thromboembolic stroke had a favorable influence on brain tissue. Chronic preventive flunarizine treatment failed to have a beneficial effect.