External Validation and Modification of the EDEMA Score for Predicting Malignant Brain Edema After Acute Ischemic Stroke

Association of Dynamic Trajectories of Time-Series Data and Life-Threatening Mass Effect in Large Middle Cerebral Artery Stroke

BACKGROUND

Life-threatening, space-occupying mass effect due to cerebral edema and/or hemorrhagic transformation is an early complication of patients with middle cerebral artery stroke. Little is known about longitudinal trajectories of laboratory and vital signs leading up to radiographic and clinical deterioration related to this mass effect.

METHODS

We curated a retrospective data set of 635 patients with large middle cerebral artery stroke totaling 95,463 data points for 10 longitudinal covariates and 40 time-independent covariates. We assessed trajectories of the 10 longitudinal variables during the 72 h preceding three outcomes representative of life-threatening mass effect: midline shift ≥ 5 mm, pineal gland shift (PGS) > 4 mm, and decompressive hemicraniectomy (DHC). We used a "backward-looking" trajectory approach. Patients were aligned based on outcome occurrence time and the trajectory of each variable was assessed before that outcome by accounting for cases and noncases, adjusting for confounders. We evaluated longitudinal trajectories with Cox proportional time-dependent regression.

RESULTS

Of 635 patients, 49.0% were female, and the mean age was 69 years. Thirty five percent of patients had midline shift ≥ 5 mm, 24.3% of patients had PGS > 4 mm, and 10.7% of patients underwent DHC. Backward-looking trajectories showed mild increases in white blood cell count (10-11 K/UL within 72 h), temperature (up to half a degree within 24 h), and sodium levels (1-3 mEq/L within 24 h) before the three outcomes of interest. We also observed a decrease in heart rate (75-65 beats per minute) 24 h before DHC. We found a significant association between increased white blood cell count with PGS > 4 mm (hazard ratio 1.05, p value 0.007).

CONCLUSIONS

Longitudinal profiling adjusted for confounders demonstrated that white blood cell count, temperature, and sodium levels appear to increase before radiographic and clinical indicators of space-occupying mass effect. These findings will inform the development of multivariable dynamic risk models to aid prediction of life-threatening, space-occupying mass effect.

Follow-up ASPECTS improves prediction of potentially lethal malignant edema in patients with large middle cerebral artery stroke

BACKGROUND

Recent studies have shown that follow-up head CT is a strong predictor of functional outcomes in patients with middle cerebral artery stroke and mechanical thrombectomy. We sought to determine whether total and/or regional follow-up Alberta Stroke Program Early CT Score (ASPECTSfu) are associated with important clinical outcomes during hospitalization and improve the performance of clinical prediction models of potentially lethal malignant edema (PLME).

METHODS

We conducted a retrospective study of patients at three medical centers in a major North American metropolitan area with baseline and follow-up head CTs after large middle cerebral artery stroke between 2006 and 2022. We used multivariable logistic regression to test the association of total and regional ASPECTSfu with PLME (cerebral edema related death or surgery), adjusting for total baseline ASPECTS, age, sex, admission glucose, tissue plasminogen activator, and mechanical thrombectomy. We compared existing clinical risk models with and without total or regional ASPECTSfu using area under the curve.

RESULTS

In our 560 patient cohort, lower total ASPECTSfu was significantly associated with higher odds of PLME when adjusting for confounders (OR 1.69, 95% CI 1.49 to 2.0), and improved model discrimination compared with existing models and models using baseline ASPECTS. Deep territory involvement (OR 2.46, 95% CI 1.53 to 4.01) and anterior territory involvement (OR 3.23, 95% CI 1.88 to 5.71) were significantly associated with PLME.

CONCLUSIONS

Lower ASPECTSfu and certain locations on regional ASPECTSfu, including deep and anterior areas, were significantly associated with PLME. Including ASPECTSfu information improved discrimination of established edema prediction models and could be used immediately to help facilitate clinical management decisions and prognostication.

HELMET: A Hybrid Machine Learning Framework for Real-Time Prediction of Edema Trajectory in Large Middle Cerebral Artery Stroke

Malignant cerebral edema occurs when brain swelling displaces and compresses vital midline structures within the first week of a large middle cerebral artery stroke. Early interventions such as hyperosmolar therapy or surgical decompression may reverse secondary injury but must be administered judiciously. To optimize treatment and reduce secondary damage, clinicians need strategies to frequently and quantitatively assess the trajectory of edema using updated, relevant information. However, existing risk assessment tools are limited by the absence of structured records capturing the evolution of edema and typically estimate risk at a single time point early in the admission, therefore failing to account for changes in variables over the following hours or days. To address this, we developed and validated dynamic machine learning models capable of accurately predicting the severity of midline structure displacement, an established indicator of malignant edema, in real-time. Our models can provide updated estimations as frequently as every hour, using data from structured time-varying patient records, radiographic text, and human-curated neurological characteristics. Our work resulted in two novel multi-class classification models, collectively named Hybrid Ensemble Learning Models for Edema Trajectory (HELMET), predicting the progression of midline shift over 8-hour (HELMET-8) and 24-hour windows (HELMET-24), respectively. HELMET combines transformer-based large language models with supervised ensemble learning, demonstrating the value of merging human expertise and multimodal health records in developing clinical risk scores. Both models were trained on a retrospective cohort of 15,696 observations from 623 patients hospitalized with large middle cerebral artery ischemic stroke and were externally validated using 3,713 observations from 60 patients at a separate hospital system. Our HELMET models are accurate and generalize effectively to diverse populations, achieving a cross-validated mean area under the receiver operating characteristic score of 96.6% in the derivation cohort and 92.5% in the external validation cohort. Moreover, our approach provides a framework for developing hybrid risk prediction models that integrate both human-extracted and algorithm-derived multi-modal inputs. Our work enables accurate estimation of complex, dynamic, and highly specific clinical targets, such as midline shift, in real-time, even when relevant structured information is limited in electronic health record databases.

Association of Serum Macrophage Migration Inhibitory Factor with 3-Month Poor Outcome and Malignant Cerebral Edema in Patients with Large Hemispheric Infarction

BACKGROUND

We aimed to investigate the associations of macrophage migration inhibitory factor (MIF), toll-like receptors 2 and 4 (TLR2/4), and matrix metalloproteinase 9 (MMP9) with 3-month poor outcome, death, and malignant cerebral edema (MCE) in patients with large hemispheric infarction (LHI).

METHODS

Patients with LHI within 24 h of onset were enrolled consecutively. Serum MIF, TLR2/4, and MMP9 concentrations on admission were measured. Poor outcome was defined as a modified Rankin Scale score of ≥ 3 at 3 months. MCE was defined as a decreased level of consciousness, anisocoria and midline shift > 5 mm or basal cistern effacement, or indications for decompressive craniectomy during hospitalization. The cutoff values for MIF/MMP9 were obtained from the receiver operating characteristic curve.

RESULTS

Of the 130 patients with LHI enrolled, 90 patients (69.2%) had 3-month poor outcome, and MCE occurred in 55 patients (42.3%). Patients with serum MIF concentrations ≤ 7.82 ng/mL for predicting 3-month poor outcome [adjusted odds ratio (OR) 2.827, 95% confidence interval (CI) 1.144-6.990, p = 0.024] also distinguished death (adjusted OR 4.329, 95% CI 1.841-10.178, p = 0.001). Similarly, MMP9 concentrations ≤ 46.56 ng/mL for predicting 3-month poor outcome (adjusted OR 2.814, 95% CI 1.236-6.406, p = 0.014) also distinguished 3-month death (adjusted OR 3.845, 95% CI 1.534-9.637, p = 0.004).

CONCLUSIONS

Lower serum MIF and MMP9 concentrations at an early stage were independently associated with 3-month poor outcomes and death in patients with LHI. These findings need further confirmation in larger sample studies.

Oral Glyburide for the Prevention of Cerebral Edema in Acute Ischemic Stroke

OBJECTIVE

The purpose of this study was to describe the impact of enteral glyburide on cerebral edema formation and hypoglycemia when used to treat patients diagnosed with acute ischemic stroke (AIS).

METHODS

This study was a single-center, retrospective medical record review that included all patients aged ≥18 years diagnosed with AIS who received ≥1 dose of enteral glyburide for the prevention of cerebral edema from January 1, 2018 to March 31, 2022. The primary outcome was the percentage of patients requiring intervention for cerebral edema management after glyburide initiation, and the safety outcome was the occurrence of hypoglycemia in this patient population.

RESULTS

The final evaluation included 44 patients, with 6 patients (14%) requiring intervention for cerebral edema after glyburide initiation. The average baseline National Institutes of Health stroke scale score was 19. Overall, in-hospital mortality was 36% (n = 17), and hypoglycemia occurred in 7 patients (15%). Of the 44 patients, 20 (45%) received a partial duration of enteral glyburide (1-4 doses) and 24 (55%) received a full duration of enteral glyburide (5-7 doses). The rate of intervention for cerebral edema (10% vs. 17%) and the incidence of hypoglycemia (5% vs. 23%) were lower in the partial duration than in the full duration group. The in-hospital all-cause mortality rate was higher in the partial duration group than in the full duration group (43% vs. 31%).

CONCLUSIONS

Despite the relatively low rates of intervention for cerebral edema, hypoglycemia was common, especially for patients receiving 5-7 doses of enteral glyburide for the prevention of cerebral edema after moderate-to-severe AIS.

Updates in Management of Large Hemispheric Infarct

This review delves into updates in management of large hemispheric infarction (LHI), a condition affecting up to 10% of patients with supratentorial strokes. While traditional management paradigms have endured, recent strides in research have revolutionized the approach to acute therapies, monitoring, and treatment. Notably, advancements in triage methodologies and the application of both pharmacological and mechanical abortive procedures have reshaped the acute care trajectory for patients with LHI. Moreover, ongoing endeavors have sought to refine strategies for the optimal surveillance and mitigation of complications, notably space-occupying mass effect, which can ensue in the aftermath of LHI. By amalgamating contemporary guidelines with cutting-edge clinical trial findings, this review offers a comprehensive exploration of the current landscape of acute and ongoing patient care for LHI, illuminating the evolving strategies that underpin effective management in this critical clinical domain.

Association of large core middle cerebral artery stroke and hemorrhagic transformation with hospitalization outcomes

Historically, investigators have not differentiated between patients with and without hemorrhagic transformation (HT) in large core ischemic stroke at risk for life-threatening mass effect (LTME) from cerebral edema. Our objective was to determine whether LTME occurs faster in those with HT compared to those without. We conducted a two-center retrospective study of patients with ≥ 1/2 MCA territory infarct between 2006 and 2021. We tested the association of time-to-LTME and HT subtype (parenchymal, petechial) using Cox regression, controlling for age, mean arterial pressure, glucose, tissue plasminogen activator, mechanical thrombectomy, National Institute of Health Stroke Scale, antiplatelets, anticoagulation, temperature, and stroke side. Secondary and exploratory outcomes included mass effect-related death, all-cause death, disposition, and decompressive hemicraniectomy. Of 840 patients, 358 (42.6%) had no HT, 403 (48.0%) patients had petechial HT, and 79 (9.4%) patients had parenchymal HT. LTME occurred in 317 (37.7%) and 100 (11.9%) had mass effect-related deaths. Parenchymal (HR 8.24, 95% CI 5.46-12.42, p < 0.01) and petechial HT (HR 2.47, 95% CI 1.92-3.17, p < 0.01) were significantly associated with time-to-LTME and mass effect-related death. Understanding different risk factors and sequelae of mass effect with and without HT is critical for informed clinical decisions.

An Explainable Artificial Intelligence Model to Predict Malignant Cerebral Edema after Acute Anterior Circulating Large-Hemisphere Infarction

INTRODUCTION

Malignant cerebral edema (MCE) is a serious complication and the main cause of poor prognosis in patients with large-hemisphere infarction (LHI). Therefore, the rapid and accurate identification of potential patients with MCE is essential for timely therapy. This study utilized an artificial intelligence-based machine learning approach to establish an interpretable model for predicting MCE in patients with LHI.

METHODS

This study included 314 patients with LHI not undergoing recanalization therapy. The patients were divided into MCE and non-MCE groups, and the eXtreme Gradient Boosting (XGBoost) model was developed. A confusion matrix was used to measure the prediction performance of the XGBoost model. We also utilized the SHapley Additive exPlanations (SHAP) method to explain the XGBoost model. Decision curve and receiver operating characteristic curve analyses were performed to evaluate the net benefits of the model.

RESULTS

MCE was observed in 121 (38.5%) of the 314 patients with LHI. The model showed excellent predictive performance, with an area under the curve of 0.916. The SHAP method revealed the top 10 predictive variables of the MCE such as ASPECTS score, NIHSS score, CS score, APACHE II score, HbA1c, AF, NLR, PLT, GCS, and age based on their importance ranking.

CONCLUSION

An interpretable predictive model can increase transparency and help doctors accurately predict the occurrence of MCE in LHI patients not undergoing recanalization therapy within 48 h of onset, providing patients with better treatment strategies and enabling optimal resource allocation.

Naoxinqing tablet protects against cerebral ischemic/reperfusion injury by regulating ampkα/NAMPT/SIRT1/PGC-1α pathway

AIM OF THE STUDY

Naoxinqing (NXQ) tablets are derived from persimmon leaves and are widely used in China for promoting blood circulation and removing blood stasis in China. We aimed to explore whether NXQ has the therapeutic effect on ischemic stroke and explored its possible mechanism.

MATERIALS AND METHODS

The cerebral artery occlusion/reperfusion (MCAO/R) surgery was used to establish the cerebral ischemic/reperfusion rat model. NXQ (60 mg/kg and 120 mg/kg) were administered orally. The TTC staining, whole brain water content, histopathology staining, immunofluorescent staining, enzyme-linked immunosorbent assay (ELISA) and Western blot analyses were performed to determine the therapeutical effect of NXQ on MCAO/R rats.

RESULTS

The study demonstrated that NXQ reduced the cerebral infarction volumes and neurologic deficits in MCAO/R rats. The neuroprotective effects of NXQ were accompanied by inhibited oxidative stress and inflammation. The nerve regeneration effects of NXQ were related to regulating the AMPKα/NAMPT/SIRT1/PGC-1α pathway.

CONCLUSION

In summary, our results revealed that NXQ had a significant protective effect on cerebral ischemia-reperfusion injury in rats. This study broadens the therapeutic scope of NXQ tablets and provides new neuroprotective mechanisms of NXQ as an anti-stroke therapeutic agent.

External validation and comparison of MBE, EDEMA, and modified EDEMA scores for predicting malignant cerebral EDEMA in Chinese patients with large hemispheric infarction patients without revascularization

BACKGROUND

Malignant cerebral edema (MCE) is a severe condition characterized by rapid neurological deterioration and a potentially poor prognosis. Scoring systems including the malignant brain edema (MBE) score, Enhanced Detection of Edema in Malignant Anterior Circulation Stroke score (EDEMA), and modified EDEMA score, have been developed to predict MCE in patients with large hemispheric infarction (LHI). We aimed to externally validate and comparethe predictive efficacy of these scores in LHI patients within 48 h of onset and not undergoing reperfusion therapy.

METHODS

Demographic, clinical and radiological data were retrospectively collected from LHI patients within 48 h of onset and not receiving reperfusion therapy. Patients were divided into MCE and non-MCE group. The calibration, discrimination, and clinical practicability of the three scores were verified using Hosmer-Lemeshow goodness-of-fit test, receiver operating characteristic (ROC) curve analysis and decision curve analysis (DCA), respectively. Finally, continuous net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were applied to determine the discrimination performance of the three scores.

RESULTS

A total of 314 patients were included in the study, with 122 cases being MCE patients. The Hosmer-Lemeshow goodness-of-fit test showed excellent fitting ability across the MBE (p = 0.36), EDEMA (p = 0.61), and modified EDEMA scores (p = 0.62) in our patients. The MBE, EDEMA, and modified EDEMA scores had the AUCs of 0.855 (95 % CI 0.818-0.898), 0.782 (95 % CI 0.727-0.837) and 0.878 (95 % CI 0.844-0.919) respectively. The MBE (NRI, 0.33; 95 % CI, 0.11-0.56, p = 0.003 and IDI, 0.11; 95 % CI, 0.03-0.18; p = 0.004) and modified EDEMA scores (NRI, 1.10; 95 % CI, 0.94-1.26; p < 0.001 and IDI, 0.17; 95 % CI, 0.13-0.20, p < 0.001) showed better performance than the EDEMA score. DCA demonstrated that the modified EDEMA score outperformed the other two scores, possessing heightened clinical usefulness.

CONCLUSIONS

The MBE, EDEMA, and modified EDEMA scores for predicting MCE are also applicable in non-revascularization LHI patients within 48 h of onset. Both the MBE and modified EDEMA scores demonstrated higher predictive validity as predictive tools for MCE in LHI patients than the EDEMA score. Furthermore, the modified EDEMA score could be a suitable prediction tool in Chinese patients for its excellent clinical utility.

Dynamic trajectories of life-threatening mass effect in patients with large middle cerebral artery stroke

Background

Life-threatening, space-occupying mass effect due to cerebral edema and/or hemorrhagic transformation is an early complication of patients with middle cerebral artery (MCA) stroke. Little is known about longitudinal trajectories of laboratory and vital signs leading up to radiographic and clinical deterioration related to this mass effect.

Methods

We curated a granular retrospective dataset of 635 patients with large middle cerebral artery (MCA) stroke totaling 108,547 data points for repeated measurements of 10 covariates, and 40 time-independent covariates. We assessed longitudinal trajectories of the 10 longitudinal variables during the 72 hours preceding three outcomes representative of life-threatening mass effect: midline shift (MLS) ≥5mm, pineal gland shift (PGS) >4mm, and decompressive hemicraniectomy (DHC). We used a "backward looking" trajectory approach. Patients were aligned according to the time of outcome occurrence and the trajectory of each variable was assessed prior to that outcome by accounting for both cases and non-cases.

Results

Of 635 patients, 49% were female, and mean age was 69 years. Thirty five percent of patients had MLS ≥ 5mm, 24.1% had PGS >4mm, and DHC occurred in 10.7%. For the three outcomes of interest, backward-looking trajectories showed mild increases in white blood cell count (10 up to 11 K/UL within 72 hours), temperature (up to half a degree within 24 hours), and sodium (1-3 mEq/L within 24 hours) leading up to outcomes. We also observed a decrease in heart rate (75 - 65 beats per minute) 24 hours prior to DHC.

Conclusions

Univariable longitudinal profiling showed that temperature, white blood cell count, and sodium increase prior to radiographic and clinical indicators of space-occupying mass effect. These findings will inform development of multivariable dynamic risk models to aid prediction of life-threatening space-occupying mass effect.

A Novel Nomogram for Predicting Malignant Cerebral Edema After Endovascular Thrombectomy in Acute Ischemic Stroke: A Retrospective Cohort Study

BACKGROUND

Malignant cerebral edema (MCE) is a common and feared complication after endovascular thrombectomy (EVT) in acute ischemic stroke (AIS). This study aimed to establish a nomogram to predict MCE in anterior circulation large vessel occlusion stroke (LVOS) patients receiving EVT in order to guide the postoperative medical care in the acute phase.

METHODS

In this retrospective cohort study, 381 patients with anterior circulation LVOS receiving EVT were screened from 636 hospitalized patients with LVOS at 2 stroke medical centers. Clinical baseline data and imaging data were collected within 2-5 days of admission to the hospital. The patients were divided into 2 groups based on whether MCE occurred after EVT. Multivariate logistic regression analysis was used to evaluate the independent risk factors for MCE and to establish a nomogram.

RESULTS

Sixty-six patients out of 381 (17.32%) developed MCE. The independent risk factors for MCE included admission National Institutes of Health Stroke Scale (NIHSS) ≥16 (odds ratio [OR] 1.851; 95% CI 1.029-3.329; P = 0.038), ASPECT score (OR 0.621; 95% CI 0.519-0.744; P < 0.001), right hemisphere (OR 1.636; 95% CI 0.941-2.843; P = 0.079), collateral circulation (OR 0.155; 95% CI 0.074-0.324; P < 0.001), recanalization (OR 0.223; 95% CI 0.109-0.457; P < 0.001), hematocrit (OR, 0.937; 95% CI: 0.892-0.985; P =0.010), and glucose (OR 1.118; 95% CI 1.023-1.223; P = 0.036), which were adopted as parameters of the nomogram. The receiver operating characteristic curve analysis showed that the area under the curve of the nomogram in predicting MCE was 0.901(95% CI 0.848-0.940; P < 0.001). The Hosmer-Lemeshow test results were not significant (P = 0.685), demonstrating a good calibration of the nomogram.

CONCLUSIONS

The novel nomogram composed of admission NIHSS, ASPECT scores, right hemisphere, collateral circulation, recanalization, hematocrit, and serum glucose provide a potential predictor for MCE in patients with AIS after EVT.

Medical management of cerebral edema in large hemispheric infarcts

Acute ischemic stroke confers a high burden of morbidity and mortality globally. Occlusion of large vessels of the anterior circulation, namely the intracranial carotid artery and middle cerebral artery, can result in large hemispheric stroke in ~8% of these patients. Edema from stroke can result in a cascade effect leading to local compression of capillary perfusion, increased stroke burden, elevated intracranial pressure, herniation and death. Mortality from large hemispheric stroke is generally high and surgical intervention may reduce mortality and improve good outcomes in select patients. For those patients who are not eligible candidates for surgical decompression either due timing, medical co-morbidities, or patient and family preferences, the mainstay of medical management for cerebral edema is hyperosmolar therapy. Other neuroprotectants for cerebral edema such as glibenclamide are under investigation. This review will discuss current guidelines and evidence for medical management of cerebral edema in large hemispheric stroke as well as discuss important neuromonitoring and critical care management targeted at reducing morbidity and mortality for these patients.

The stress hyperglycemia ratio is associated with the development of cerebral edema and poor functional outcome in patients with acute cerebral infarction

Background and purpose

Absolute hyperglycemia at admission has been shown to be associated with the development of cerebral edema (CED) after acute cerebral infarction. Stress hyperglycemia is a more objective reflection of hyperglycemic state than absolute hyperglycemia. However, studies on the associations between stress hyperglycemia and CED are limited. We aimed to explore the associations of stress hyperglycemia, measured by stress hyperglycemia ratio (SHR), with the development of CED and poor functional outcome of acute cerebral infarction.

Methods

Patients with acute middle artery cerebral infarction admitted to the Department of Neurology, West China Hospital of Sichuan University, within 24 h of symptom onset from January 2017 to March 2021 were included. Stress hyperglycemia was assessed by the SHR: admission fasting plasma glucose (FPG)/hemoglobin A1c (HbA1c). The primary outcome was the degree of CED evaluated on brain image. The secondary outcomes were moderate-to-severe CED, poor functional outcome (modified Rankin Scale score &gt; 2), and death at 90 days. The associations between the SHR and outcomes were assessed with multivariate logistic regression analyses. We further compared the predictive value of the SHR, admission random plasma glucose (RPG), and admission FPG for outcomes in the training dataset and validation dataset.

Results

638 patients were enrolled. Each 0.1-point increase in the SHR was independently associated with a 1.31-fold increased risk of a higher degree of CED [odds ratio (OR): 1.31 (95% confidence interval (CI): 1.20–1.42), P &lt; 0.001]. The SHR was independently associated with moderate-to-severe CED [per 0.1-point increase: OR: 1.39 (95% CI: 1.24–1.57), P &lt; 0.001], poor functional outcome [per 0.1-point increase: OR: 1.25 (95% CI: 1.12–1.40), P &lt; 0.001], and death [per 0.1-point increase: OR: 1.13 (95% CI: 1.03–1.25), P &lt; 0.05]. The predictive value of the SHR (as a continuous variable), exhibited by the area under the curve in receiver operating characteristic analysis, was higher than that of the RPG and FPG for moderate-to-severe CED and poor functional outcome (P &lt; 0.05).

Conclusion

The SHR is independently associated with the severity of CED, poor functional outcome, and death after acute cerebral infarction, and the SHR (as a continuous variable) has a better predictive value for moderate-to-severe CED and poor functional outcome than the RPG and FPG.

Predictors and Dynamic Nomogram to Determine the Individual Risk of Malignant Brain Edema After Endovascular Thrombectomy in Acute Ischemic Stroke

Background and Purpose

This study aimed to construct an optimal dynamic nomogram for predicting malignant brain edema (MBE) in acute ischemic stroke (AIS) patients after endovascular thrombectomy (ET).

Methods

We enrolled AIS patients after ET from May 2017 to April 2021. MBE was defined as a midline shift of >5 mm at the septum pellucidum or pineal gland based on follow-up computed tomography within 5 days after ET. Multivariate logistic regression and LASSO (least absolute shrinkage and selection operator) regression were used to construct the nomogram. The area under the receiver operating characteristic curve (AUC) and decision-curve analysis were used to compare our nomogram with two previous risk models for predicting brain edema after ET.

Results

MBE developed in 72 (21.9%) of the 329 eligible patients. Our dynamic web-based nomogram (https://successful.shinyapps.io/DynNomapp/) consisted of five parameters: basal cistern effacement, postoperative National Institutes of Health Stroke Scale (NIHSS) score, brain atrophy, hypoattenuation area, and stroke etiology. The nomogram showed good discrimination ability, with a C-index (Harrell’s concordance index) of 0.925 (95% confidence interval=0.890–0.961), and good calibration (Hosmer-Lemeshow test, p=0.386). All variables had variance inflation factors of <1.5 and tolerances of >0.7, suggesting no significant collinearity among them. The AUC of our nomogram (0.925) was superior to those of Xiang-liang Chen and colleagues (0.843) and Ming-yang Du and colleagues (0.728).

Conclusions

Our web-based dynamic nomogram reliably predicted the risk of MBE in AIS patients after ET, and hence is worthy of further evaluation.

Automated quantitative lesion water uptake in acute stroke is a predictor of malignant cerebral edema

OBJECTIVES

Net water uptake (NWU) has been shown to have a linear relationship with brain edema. Based on an automated-Alberta Stroke Program Early Computed Tomography Score (ASPECTS) technique, we automatically derived NWU from baseline multimodal computed tomography (CT), namely ASPECTS-NWU. We aimed to determine if ASPECTS-NWU can predict the development of malignant cerebral edema (MCE).

METHODS

One hundred and forty-six patients with large-vessel occlusion were retrospectively enrolled. Quantitative NWU based on automated-ASPECTS was measured both on nonenhanced CT (NECT) and CT angiography (CTA), namely NECT-ASPECT-NWU and CTA-ASPECTS-NWU. The correlation between ASPECTS-NWU and cerebral edema (CED) grades was calculated using Spearman rank correlation. Univariate logistic regression was used to assess the effect of radiological and clinical features on MCE, and a multivariable model with significant factors from the univariate regression analysis was built. Receiver operating characteristic (ROC) was obtained and area under curve (AUC) was compared.

RESULTS

CTA-ASPECTS-NWU had a moderate positive correlation with CED grades (r = 0.62; 95% confidence interval [CI], 0.51-0.71; p < 0.001). The CTA-ASPECTS-NWU performed better than the NECT-ASPECTS-NWU with AUC: 0.88 vs. 0.71 (p < 0.001). Multivariable logistic regression model integrating CTA-ASPECTS-NWU, collateral score, and age showed the CTA-ASPECTS-NWU was an independent predictor of MCE with an AUC of 0.94 (95% CI: 0.90-0.98; p < 0.001).

CONCLUSIONS

This study demonstrates that ASPECTS-NWU is a quantitative predictor of MCE after large-vessel occlusion of the middle cerebral artery territory. The multivariable logistic regression model may enhance the identification of patients with MCE needing anti-edematous treatment.

KEY POINTS

• The automated-ASPECTS technique can automatically detect the affected regions with early ischemic changes and NWU could be manually calculated. • The CTA-ASPECTS-NWU performs better than the NECT-ASPECTS-NWU on predicting the development of MCE. • The multivariable logistic regression model may enhance the identification of patients with MCE needing anti-edematous treatment.

Accelerating Prediction of Malignant Cerebral Edema After Ischemic Stroke with Automated Image Analysis and Explainable Neural Networks

BACKGROUND

Malignant cerebral edema is a devastating complication of stroke, resulting in deterioration and death if hemicraniectomy is not performed prior to herniation. Current approaches for predicting this relatively rare complication often require advanced imaging and still suffer from suboptimal performance. We performed a pilot study to evaluate whether neural networks incorporating data extracted from routine computed tomography (CT) imaging could enhance prediction of edema in a large diverse stroke cohort.

METHODS

An automated imaging pipeline retrospectively extracted volumetric data, including cerebrospinal fluid (CSF) volumes and the hemispheric CSF volume ratio, from baseline and 24 h CT scans performed in participants of an international stroke cohort study. Fully connected and long short-term memory (LSTM) neural networks were trained using serial clinical and imaging data to predict those who would require hemicraniectomy or die with midline shift. The performance of these models was tested, in comparison with regression models and the Enhanced Detection of Edema in Malignant Anterior Circulation Stroke (EDEMA) score, using cross-validation to construct precision-recall curves.

RESULTS

Twenty of 598 patients developed malignant edema (12 required surgery, 8 died). The regression model provided 95% recall but only 32% precision (area under the precision-recall curve [AUPRC] 0.74), similar to the EDEMA score (precision 28%, AUPRC 0.66). The fully connected network did not perform better (precision 33%, AUPRC 0.71), but the LSTM model provided 100% recall and 87% precision (AUPRC 0.97) in the overall cohort and the subgroup with a National Institutes of Health Stroke Scale (NIHSS) score ≥ 8 (p = 0.0001 vs. regression and fully connected models). Features providing the most predictive importance were the hemispheric CSF ratio and NIHSS score measured at 24 h.

CONCLUSIONS

An LSTM neural network incorporating volumetric data extracted from routine CT scans identified all cases of malignant cerebral edema by 24 h after stroke, with significantly fewer false positives than a fully connected neural network, regression model, and the validated EDEMA score. This preliminary work requires prospective validation but provides proof of principle that a deep learning framework could assist in selecting patients for surgery prior to deterioration.

Computed Tomography Based Score of Early Ischemic Changes Predicts Malignant Infarction

Background and Purpose: Identification of ischemic stroke patients at high risk of developing life-threatening malignant infarction at an early stage is critical to consider more rigorous monitoring and further therapeutic measures. We hypothesized that a score consisting of simple measurements of visually evident ischemic changes in non-enhanced CT (NEMMI score) predicts malignant middle cerebral artery (MCA) infarctions (MMI) with similar diagnostic power compared to other baseline clinical and imaging parameters. Methods: One hundred and nine patients with acute proximal MCA occlusion were included. Fifteen (13.8%) patients developed MMI. NEMMI score was defined using the sum of the maximum diameter (anterior-posterior plus medio-lateral) of the hypoattenuated lesion in baseline-CT multiplied by a hypoattenuation factor (3-point visual grading in non-enhanced CT, no/subtle/clear hypoattenuation = 1/2/3). Receiver operating characteristic (ROC) curve analysis and multivariable logistic regression analysis were used to calculate the predictive values of the NEMMI score, baseline clinical and other imaging parameters. Results: The median NEMMI score at baseline was 13.6 (IQR: 11.6-31.1) for MMI patients, and 7.7 (IQR: 3.9-11.2) for patients with non-malignant infarctions (p < 0.0001). Based on ROC curve analysis, a NEMMI score >10.5 identified MMI with good discriminative power (AUC: 0.84, sensitivity/specificity: 93.3/70.7%), which was higher compared to age (AUC: 0.76), NIHSS (AUC: 0.61), or ischemic core volume (AUC: 0.80). In multivariable logistic regression analysis, NEMMI score was significantly and independently associated with MMI (OR: 1.33, 95%CI: 1.13-1.56, p < 0.001), adjusted for recanalization status. Conclusion: The NEMMI score is a quick and simple rating tool of early ischemic changes on CT and could serve as an important surrogate marker for developing malignant edema. Its diagnostic accuracy was similar to CTP and clinical parameters.

Absent Cortical Venous Filling Is Associated with Aggravated Brain Edema in Acute Ischemic Stroke

BACKGROUND AND PURPOSE

Predicting malignant cerebral edema can help identify patients who may benefit from appropriate evidence-based interventions. We investigated whether absent cortical venous filling is associated with more pronounced early brain edema, which leads to malignant cerebral edema.

MATERIALS AND METHODS

Patients with acute ischemic stroke caused by large-vessel occlusion in the MCA territory who presented between July 2017 and September 2019 to our hospital were included. Collateral filling was rated using the modified Tan scale on CTA, and good collaterals were defined as a score of 2-3. The Cortical Vein Opacification Score (COVES) was calculated, and absent cortical venous filling was defined as a score of 0. Early brain edema was determined using net water uptake on baseline CT images. Malignant cerebral edema was defined as a midline shift of ≥5 mm on follow-up imaging or a massive cerebral swelling leading to decompressive hemicraniectomy or death. Multivariate linear and logistic regression models were performed to analyze data.

RESULTS

A total of 163 patients were included. Net water uptake was significantly higher in patients with absent than in those with favorable cortical venous filling (8.1% versus 4.2%; P < .001). In the multivariable regression analysis, absent cortical venous filling (β = 2.04; 95% CI, 0.75-3.32; P = .002) was significantly and independently associated with higher net water uptake. Absent cortical venous filling (OR, 14.68; 95% CI, 4.03-53.45; P < .001) and higher net water uptake (OR, 1.29; 95% CI, 1.05-1.58; P = .016) were significantly associated with increased likelihood of malignant cerebral edema.

CONCLUSIONS

Patients with absent cortical venous filling were associated with an increased early brain edema and a higher risk of malignant cerebral edema. These patients may be targeted for optimized adjuvant antiedematous treatment.

Malignant infarction after endovascular treatment: Incidence and prediction

BACKGROUND

Early prediction of malignant infarction may guide treatment decisions. For patients who received endovascular treatment, the risk of malignant infarction is unknown and risk factors are unrevealed.

AIMS

The objective of this study is to estimate the incidence of malignant infarction after endovascular treatment in patients with an occlusion of the anterior circulation, to identify independent risk factors, and to establish a model for prediction.

METHODS

We analyzed patients who received endovascular treatment for a large vessel occlusion in the anterior circulation within 6.5 h after symptom onset, included in the Dutch MR CLEAN Registry between March 2014 and June 2016. We compared patients with and without malignant infarction. Candidate predictors were incorporated in a multivariable binary logistic regression model. The final prediction model was established using backward elimination. Discrimination and calibration were evaluated with the area under the receiver operating characteristic curve (AUROC) and the Hosmer-Lemeshow test.

RESULTS

Of 1445 patients, 82 (6%) developed malignant infarction. Independent predictors were lower age, higher National Institutes of Health Stroke Scale (NIHSS), lower alberta stroke program early CT score (ASPECTS), internal carotid artery occlusion, lower collateral score, longer times from onset to groin puncture, and unsuccessful reperfusion. The AUROC of a prediction model combining these features was 0.83 (95% confidence interval (CI): 0.79-0.88) and the Hosmer-Lemeshow test indicated appropriate calibration (P = 0.937).

CONCLUSION

The risk of malignant infarction after endovascular treatment started within 6.5 h of stroke onset is approximately 6%. Successful reperfusion decreases the risk. A prediction model combining easily retrievable measures of age, ASPECTS, collateral status, and reperfusion shows good discrimination between patients who will develop malignant infarction and those who will not.

Image Patch-Based Net Water Uptake and Radiomics Models Predict Malignant Cerebral Edema After Ischemic Stroke

Malignant cerebral edema (MCE) after an ischemic stroke results in a poor outcome or death. Early prediction of MCE helps to identify subjects that could benefit from a surgical decompressive craniectomy. Net water uptake (NWU) in an ischemic lesion is a predictor of MCE; however, CT perfusion and lesion segmentation are required. This paper proposes a new Image Patch-based Net Water Uptake (IP-NWU) procedure that only uses non-enhanced admission CT and does not need lesion segmentation. IP-NWU is calculated by comparing the density of ischemic and contralateral normal patches selected from the middle cerebral artery (MCA) area using standard reference images. We also compared IP-NWU with the Segmented Region-based NWU (SR-NWU) procedure in which segmented ischemic regions from follow-up CT images are overlaid onto admission images. Furthermore, IP-NWU and its combination with imaging features are used to construct predictive models of MCE with a radiomics approach. In total, 116 patients with an MCA infarction (39 with MCE and 77 without MCE) were included in the study. IP-NWU was significantly higher for patients with MCE than those without MCE (p < 0.05). IP-NWU can predict MCE with an AUC of 0.86. There was no significant difference between IP-NWU and SR-NWU, nor between their predictive efficacy for MCE. The inter-reader and interoperation agreement of IP-NWU was exceptional according to the Intraclass Correlation Coefficient (ICC) analysis (inter-reader: ICC = 0.92; interoperation: ICC = 0.95). By combining IP-NWU with imaging features through a random forest classifier, the radiomics model achieved the highest AUC (0.96). In summary, IP-NWU and radiomics models that combine IP-NWU with imaging features can precisely predict MCE using only admission non-enhanced CT images scanned within 24 h from onset.

Prediction of Malignant Acute Middle Cerebral Artery Infarction via Computed Tomography Radiomics

Malignant middle cerebral artery infarction (mMCAi) is a serious complication of cerebral infarction usually associated with poor patient prognosis. In this retrospective study, we analyzed clinical information as well as non-contrast computed tomography (NCCT) and computed tomography angiography (CTA) data from patients with cerebral infarction in the middle cerebral artery (MCA) territory acquired within 24 h from symptoms onset. Then, we aimed to develop a model based on the radiomics signature to predict the development of mMCAi in cerebral infarction patients. Patients were divided randomly into training (n = 87) and validation (n = 39) sets. A total of 396 texture features were extracted from each NCCT image from the 126 patients. The least absolute shrinkage and selection operator regression analysis was used to reduce the feature dimension and construct an accurate radiomics signature based on the remaining texture features. Subsequently, we developed a model based on the radiomics signature and Alberta Stroke Program Early CT Score (ASPECTS) based on NCCT to predict mMCAi. Our prediction model showed a good predictive performance with an AUC of 0.917 [95% confidence interval (CI), 0.863-0.972] and 0.913 [95% CI, 0.795-1] in the training and validation sets, respectively. Additionally, the decision curve analysis (DCA) validated the clinical efficacy of the combined risk factors of radiomics signature and ASPECTS based on NCCT in the prediction of mMCAi development in patients with acute stroke across a wide range of threshold probabilities. Our research indicates that radiomics signature can be an instrumental tool to predict the risk of mMCAi.