External Validation of Five Scores to Predict Stroke-Associated Pneumonia and the Role of Selected Blood Biomarkers

The Trends of Neutrophil-to-Lymphocyte Ratio, Platelet-to-Lymphocyte Ratio, and Systemic Immunoinflammatory Index in Patients with Intracerebral Hemorrhage and Clinical Value in Predicting Pneumonia 30 Days After Surgery

BACKGROUND

Inflammatory response is closely associated with secondary brain injury and pneumonia in intracerebral hemorrhage (ICH). In this study, we aimed to investigate the value of neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and systemic immunoinflammatory index (SII) in the development of pneumonia in ICH patients 30 days after surgery.

METHODS

We retrospectively collected clinical data on patients with ICH who underwent surgical treatment at our institution from January 2016 to December 2022, mainly including NLR, PLR, and SII at different time points. Receiver operating characteristic curves were used to compare the value of different inflammatory indicators in predicting the development of postoperative pneumonia 30 days after surgery in ICH patients, and multivariate logistic regression analyses were used to identify independent risk factors for pneumonia 30 days after surgery.

RESULTS

Among 112 patients with ICH undergoing surgical treatment, 31 (27.7%) developed pneumonia postoperatively. The results of the univariate analysis demonstrated that patients in the pneumonia group experienced significantly higher blood glucose, NLR at 72 hours postoperatively, PLR at 72 hours postoperatively, and SII at 72 hours postoperatively (SII3) than those in the nonpneumonia group, and significantly lower admission Glasgow Coma Scale scores than those in the nonpneumonia group (all P < 0.05). NLR, PLR, and SII showed increasing and then decreasing in the disease process of ICH and peaked at 48 hours postoperatively. Multivariable logistic regression analysis revealed that SII3 was an independent risk factor for postoperative pneumonia 30 days after surgery in ICH patients (odds ratio = 1.001, 95% confidence interval: 1.000-1.002, P = 0.008). The area under the curve of the developed nomogram model was 0.895 (95% confidence interval = 0.823-0.967), with a sensitivity and specificity of 0.903 and 0.815, respectively, providing good predictive power.

CONCLUSIONS

In the course of ICH, NLR, PLR, and SII increased and then decreased and peaked at 48 hours postoperatively. The SII3 was the best predictor of the occurrence of pneumonia postoperatively in ICH patients.

Association Between TCBI (Triglycerides, Total Cholesterol, and Body Weight Index) and Stroke-Associated Pneumonia in Acute Ischemic Stroke Patients

Purpose

Stroke-associated pneumonia (SAP) usually complicates stroke and is linked to adverse prognoses. Triglycerides, total cholesterol, and body weight index (TCBI) is a new and simple calculated nutrition index. This study seeks to investigate the association between TCBI and SAP incidence, along with its predictive value.

Patients and Methods

Nine hundred and sixty-two patients with acute ischemic stroke were divided into SAP group and Non-SAP group. The TCBI was divided into three layers: T1, TCBI < 948.33; T2, TCBI 948.33-1647.15; T3, TCBI > 1647.15. Binary Logistic regression analysis was used to determine the relationship between TCBI levels and the incidence of SAP. Furthermore, restricted cubic splines (RCS) analysis was utilized to evaluate the influence of TCBI on the risk of SAP.

Results

TCBI in the SAP group was markedly lower compared to that in the Non-SAP group (P < 0.001). The Logistic regression model revealed that, using T3 layer as the reference, T1 layer had the highest risk for SAP prevalence (OR = 2.962, 95% CI: 1.600-5.485, P = 0.001), with confounding factors being controlled. The RCS model found that TCBI had a linear relationship with SAP (P for nonlinear = 0.490, P for overall = 0.004). Moreover, incorporating TCBI into the A2DS2 (Age, atrial fibrillation, dysphagia, sex, and severity) model substantially enhanced the initial model's predictive accuracy.

Conclusion

Low TCBI was associated with a higher risk of SAP. In clinical practice, TCBI has shown predictive value for SAP, contributing to early intervention and treatment of SAP.

Correlation of LP-PLA2 and MMP-9 with the occurrence of early neurological deterioration in patients with acute ischemic stroke

Early neurological deterioration is a common complication of acute ischemic stroke (AIS), which aggravates symptoms, worsens the condition, and counteracts the benefits of clinical treatment. The aim of this paper was to analyze the correlation between lipoprotein-associated phospholipase A2 (Lp-PLA2), matrix metalloproteinase-9 (MMP-9), and the occurrence of early neurological deterioration (END) in patients with AIS and to explore the clinical prediction of END by the combination of the 2 assays for the clinical prediction of END. A total of 500 AIS patients admitted to our hospital from October 2022 to October 2023 were included as study subjects, and the clinical data of all AIS patients were collected and organized to detect the levels of Lp-PLA2 and MMP-9. Categorized into END and non-END groups according to whether END occurred within 7 days of the onset of AIS, and comparing the clinical baseline data and laboratory index levels of the 2 groups. Logistic regression analysis was performed to determine the independent predictors of END, and the predictive effects of Lp-PLA2 and MMP-9 levels on END were assessed by subject work characteristics (ROC) curves. END occurred in 111 (22.2%) of 500 AIS patients. Multivariate logistic regression analysis showed that diabetes (OR 2.717, 95% CI:1.53-4.81, P < .001), baseline NIHSS score (OR 1.65, 95% CI:1.41-1.94, P < .001), Lp-PLA2 (OR 1.07, 95% CI:1.05-1.09, P < .001) and MMP-9 (OR 1.12, 95% CI:1.09-1.16, P < .001) levels were independent influences on the occurrence of END in patients with AIS after correcting for confounders. ROC curve analysis showed that Lp-PLA2, MMP-9, and a combination of both predicted END with an area under the curve was 0.730, 0.763, and 0.831, respectively, and the area under the curve for the combination of both predicting END was significantly higher than that for any of the inflammatory markers alone (P < .05). Both inflammatory markers, Lp-PLA2 and MMP-9, were independent predictors of the development of END in patients with AIS, and the combination of the two had a higher predictive value.

Magnetic resonance imaging-based deep learning imaging biomarker for predicting functional outcomes after acute ischemic stroke

PURPOSE

Clinical risk scores are essential for predicting outcomes in stroke patients. The advancements in deep learning (DL) techniques provide opportunities to develop prediction applications using magnetic resonance (MR) images. We aimed to develop an MR-based DL imaging biomarker for predicting outcomes in acute ischemic stroke (AIS) and evaluate its additional benefit to current risk scores.

METHOD

This study included 3338 AIS patients. We trained a DL model using deep neural network architectures on MR images and radiomics to predict poor functional outcomes at three months post-stroke. The DL model generated a DL score, which served as the DL imaging biomarker. We compared the predictive performance of this biomarker to five risk scores on a holdout test set. Additionally, we assessed whether incorporating the imaging biomarker into the risk scores improved the predictive performance.

RESULTS

The DL imaging biomarker achieved an area under the receiver operating characteristic curve (AUC) of 0.788. The AUCs of the five studied risk scores were 0.789, 0.793, 0.804, 0.810, and 0.826, respectively. The imaging biomarker's predictive performance was comparable to four of the risk scores but inferior to one (p = 0.038). Adding the imaging biomarker to the risk scores improved the AUCs (p-values) to 0.831 (0.003), 0.825 (0.001), 0.834 (0.003), 0.836 (0.003), and 0.839 (0.177), respectively. The net reclassification improvement and integrated discrimination improvement indices also showed significant improvements (all p < 0.001).

CONCLUSIONS

Using DL techniques to create an MR-based imaging biomarker is feasible and enhances the predictive ability of current risk scores.

Radiomic Features of Acute Cerebral Hemorrhage on Non-Contrast CT Associated with Patient Survival

The mortality rate of acute intracerebral hemorrhage (ICH) can reach up to 40%. Although the radiomics of ICH have been linked to hematoma expansion and outcomes, no research to date has explored their correlation with mortality. In this study, we determined the admission non-contrast head CT radiomic correlates of survival in supratentorial ICH, using the Antihypertensive Treatment of Acute Cerebral Hemorrhage II (ATACH-II) trial dataset. We extracted 107 original radiomic features from n = 871 admission non-contrast head CT scans. The Cox Proportional Hazards model, Kaplan-Meier Analysis, and logistic regression were used to analyze survival. In our analysis, the "first-order energy" radiomics feature, a metric that quantifies the sum of squared voxel intensities within a region of interest in medical images, emerged as an independent predictor of higher mortality risk (Hazard Ratio of 1.64, p < 0.0001), alongside age, National Institutes of Health Stroke Scale (NIHSS), and baseline International Normalized Ratio (INR). Using a Receiver Operating Characteristic (ROC) analysis, "the first-order energy" was a predictor of mortality at 1-week, 1-month, and 3-month post-ICH (all p < 0.0001), with Area Under the Curves (AUC) of >0.67. Our findings highlight the potential role of admission CT radiomics in predicting ICH survival, specifically, a higher "first-order energy" or very bright hematomas are associated with worse survival outcomes.

Machine learning using multimodal and autonomic nervous system parameters predicts clinically apparent stroke-associated pneumonia in a development and testing study

BACKGROUND

Stroke-associated pneumonia (SAP) is a preventable determinant for poor outcome after stroke. Machine learning (ML) using large-scale clinical data warehouses may be able to predict SAP and identify patients for targeted interventions. The aim of this study was to develop a prediction model for identifying clinically apparent SAP using automated ML.

METHODS

The ML model used clinical and laboratory parameters along with heart rate (HR), heart rate variability (HRV), and blood pressure (BP) values obtained during the first 48 h after stroke unit admission. A logistic regression classifier was developed and internally validated with a nested-cross-validation (nCV) approach. For every shuffle, the model was first trained and validated with a fixed threshold for 0.9 sensitivity, then finally tested on the out-of-sample data and benchmarked against a widely validated clinical score (A2DS2).

RESULTS

We identified 2390 eligible patients admitted to two-stroke units at Charité between October 2020 and June 2023, of whom 1755 had all parameters available. SAP was diagnosed in 96/1755 (5.5%). Circadian profiles in HR, HRV, and BP metrics during the first 48 h after admission exhibited distinct differences between patients with SAP diagnosis vs. those without. CRP, mRS at admission, leukocyte count, high-frequency power in HRV, stroke severity at admission, sex, and diastolic BP were identified as the most informative ML features. We obtained an AUC of 0.91 (CI 0.88-0.95) for the ML model on the out-of-sample data in comparison to an AUC of 0.84 (CI 0.76-0.91) for the previously established A2DS2 score (p < 0.001). The ML model provided a sensitivity of 0.87 (CI 0.75-0.97) with a corresponding specificity of 0.82 (CI 0.78-0.85) which outperformed the A2DS2 score for multiple cutoffs.

CONCLUSIONS

Automated, data warehouse-based prediction of clinically apparent SAP in the stroke unit setting is feasible, benefits from the inclusion of vital signs, and could be useful for identifying high-risk patients or prophylactic pneumonia management in clinical routine.

Construction and comparison of short-term prognosis prediction model based on machine learning in acute ischemic stroke

Objective

To construct and compared the short-term prognosis prediction models of acute ischemic stroke (AIS) by machine learning (ML).

Methods

Retrospectively study. The group W (mRS≤3) was clustered, and combined with group P (mRS>3) to form the post-clustering dataset for modeling. The "glmnet", "rpart", "xgboost", "randomForest", "neuralnet" packages were used to construct ML models. The accuracy, sensitivity, specificity, positive predict value (PPV), negative predict value (NPV) among the models were compared. Four external clinical datasets were used for external clinical validation. The optimal prediction model was determined by variable screening ability, model visualization, and external clinical validation performance.

Results

The post-clustering dataset contains 139 patients (group W) and 122 patients (group P). The neutrophil multiplied by D-dimer (NDM) has predictive value in all ML prediction models in this study. In the decision tree model, NDMQ occupies the first tree node, When NDM≤5.62 and the age<74.5, the probability of poor prognosis of AIS is less than 20 %. When NDM>5.62 and accompanied by pneumonia, the incidence of poor prognosis of AIS is about 90 %. In the Random Forest (RF) model, NDMQ had the highest Gini index. The variable combination screened by the RF model had the best performance in the neural network, and the accuracy, sensitivity, specificity, PPV, and NPV of the external validation were 0.800, 0.774, 0.833, 0.857, and 0.741, respectively. The RF model had the best performance in the external clinical validation datasets, with accuracies of 0.646, 0.697, 0.695, and 0.713, respectively.

Conclusions

NDM shows predictive value for AIS short-term prognosis in all ML models in this study. The optimal model in screening characteristic variables and the performance of in external clinical datasets was RF model. In the analysis of medical data with small sample size and outcome as categorical variables, RF could be used as the main algorithm to build a model.

Clinical narratives as a predictor for prognosticating functional outcomes after intracerebral hemorrhage

BACKGROUND

Intracerebral hemorrhage (ICH) is a devastating stroke type that causes high mortality rates and severe disability among survivors. Many prognostic models are available for prognosticating patients with ICH. This study aimed to investigate whether clinical narratives can improve the performance for predicting functional outcomes after ICH.

METHODS

This study used data from the hospital stroke registry and electronic health records. The study population (n = 1363) was randomly divided into a training set (75%, n = 1023) and a holdout test set (25%, n = 340). Five risk scores for ICH were used as baseline prognostic models. Using natural language processing (NLP), text-based markers were generated from the clinical narratives of the training set through machine learning (ML) and deep learning (DL) approaches. The primary outcome was a poor functional outcome (modified Rankin Scale score of 3 to 6) at hospital discharge. The predictive performance was compared between the baseline models and models enhanced by incorporating the text-based markers using the holdout test set.

RESULTS

The enhanced prognostic models outperformed the baseline models, regardless of whether ML or DL approaches were used. The areas under the receiver operating characteristic curve (AUCs) of the baseline models were between 0.760 and 0.892. Adding the text-based marker to the baseline models significantly increased the model discrimination, with AUCs ranging from 0.861 to 0.914. The net reclassification improvement and integrated discrimination improvement indices also showed significant improvements.

CONCLUSIONS

Using NLP to extract textual information from clinical narratives could improve the predictive performance of all baseline prognostic models for ICH.

Predictors of stroke-associated pneumonia and the predictive value of neutrophil percentage-to-albumin ratio

BACKGROUND

Early recognition of stroke-associated pneumonia (SAP) is critical to reducing morbidity and mortality associated with SAP. This study investigated the predictors of SAP, and the predictive value of the neutrophil percentage-to-albumin ratio (NPAR) for SAP.

METHODS

This retrospective cohort study was conducted among stroke patients admitted to Jordan University Hospital from January 2015 to May 2021. Multivariable logistic regression was used to identify independent predictors for SAP. The predictive performance was assessed using C-statistics, described as the area under the receiver-operating characteristic curve (AUC, ROC) with a 95% confidence interval.

RESULTS

Four hundred and six patients were included in the analysis, and the prevalence of SAP was 19.7%. Multivariable logistic analysis showed that males (Adjusted Odds Ratio (AOR): 5.74; 95% Confidence Interval (95%CI): 2.04-1 6.1)], dysphagia (AOR: 5.29; 95% CI: 1.80-15.5), hemiparesis (AOR: 3.27; 95% CI: 1.13-9.47), lower GCS score (AOR: 0.73; 95% CI: 0.58-0.91), higher levels of neutrophil-lymphocyte ratio (NLR) (AOR: 1.15; 95% CI: 1.07-1.24), monocyte-lymphocyte ratio (MLR) (AOR: 1.49; 95% CI: 1.13-1.96), and neutrophil percentage to albumin ratio (NPAR) (AOR: 1.53; 95% CI: 1.33-1.76) were independent predictors of SAP. The NPAR demonstrated a significantly higher AUC than both the NLR (0.939 versus 0.865, Z = 3.169, p = 0.002) and MLR (0.939 versus 0.842, Z = 3.940, p < 0.001). The AUCs of the NLR and MLR were comparable (0.865 versus 0.842, Z = 1.274, p = 0.203).

CONCLUSION

Male gender, dysphagia and hemiparesis were the strongest predictors of SAP, and NPAR has an excellent performance in predicting SAP which was better than high NLR and MLR.

Association between malnutrition and stroke-associated pneumonia in patients with ischemic stroke

BACKGROUND

Malnutrition is associated with a high risk of mortality in adults with ischemic stroke (IS). This study aimed to investigate the relationship between malnutrition and the risk of stroke-associated pneumonia (SAP) as only a few studies examined the relationship between malnutrition and the risk of SAP in IS.

METHODS

Patients were included from emergency departments of five tertiary hospitals in the REtrospective Multicenter study for Ischemic Stroke Evaluation (REMISE) study from January 2020 to December 2020. Malnutrition was defined according to the Controlling Nutritional Status (CONUT), Geriatric Nutritional Risk Index (GNRI), and Prognostic Nutritional Index (PNI) systems. Multivariable logistic regression analysis was used to explore the association between malnutrition and risk of SAP.

RESULTS

We enrolled 915 patients with IS, 193 (14.75%), 495 (54.1%), and 148 (16.2%) of whom were malnourished according to the PNI, CONUT, and GNRI scores, respectively. SAP occurred in 294 (32.1%) patients. After adjusting for confounding influencing factors in the logistic regression analysis, malnutrition (moderate and severe risk vs. absent malnutrition) was independently associated with an increased risk of SAP based on the PNI (odds ratio [OR], 5.038; 95% confidence interval [CI] 2.435-10.421, P < 0.001), CONUT (OR, 6.941; 95% CI 3.034-15.878, P < 0.001), and GNRI (OR, 2.007; 95% CI 1.186-3.119, P = 0.005) scores. Furthermore, adding malnutrition assessment indices to the A2DS2 score significantly improved the ability to predict SAP by analysis of receiver operating characteristic curves and net reclassification improvement.

CONCLUSION

Malnutrition was notably prevalent in patients with IS and independently associated with an increased risk of SAP. Further studies are required to identify the effect of interventions on malnutrition to reduce the risk of SAP.

Procalcitonin as a biomarker of nosocomial pneumonia in aneurysmal subarachnoid hemorrhage patients treated in neuro-ICU

BACKGROUND

Nosocomial pneumonia commonly develops in aneurysmal subarachnoid hemorrhage (aSAH) patients and is associated with poor prognosis of these patients. This study is designed to verify the predictive value of procalcitonin (PCT) on nosocomial pneumonia in aSAH patients.

METHODS

298 aSAH patients received treatments in the neuro-intensive care unit (NICU) of West China hospital were included. Logistic regression was conducted to verify the association between PCT level and nosocomial pneumonia and to construct a model for predicting pneumonia. Area under the receiver operating characteristic curve (AUC) were calculated to evaluate the accuracy of the single PCT and the constructed model.

RESULTS

90 (30.2%) patients developed pneumonia during hospitalizations among included aSAH patients. Pneumonia group had higher procalcitonin level (p < 0.001) than non-pneumonia group. The mortality (p < 0.001), mRS (p < 0.001), length of ICU stay (p < 0.001), length of hospital stay (p < 0.001) were both higher or longer in pneumonia group. Multivariate logistic regression indicated WFNS (p = 0.001), acute hydrocephalus (p = 0.007), WBC (p = 0.021), PCT (p = 0.046) and C-reactive protein (CRP) (p = 0.031) were independently associated with the development of pneumonia in included patients. The AUC value of procalcitonin for predicting nosocomial pneumonia was 0.764. Composed of WFNS, acute hydrocephalus, WBC, PCT and CRP, the predictive model for pneumonia has higher AUC of 0.811.

CONCLUSIONS

PCT is an available and effective predictive marker of nosocomial pneumonia in aSAH patients. Composed of WFNS, acute hydrocephalus, WBC, PCT and CRP, our constructed predictive model is helpful for clinicians to evaluate the risk of nosocomial pneumonia and guide therapeutics in aSAH patients.

Evaluation of a novel scoring system based on thrombosis and inflammation for predicting stroke-associated pneumonia: A retrospective cohort study

Background

Inflammation and thrombosis are involved in the development of stroke-associated pneumonia (SAP). Our aim was to evaluate the predictive value of a novel, simplified, thrombo-inflammatory prognostic score (TIPS) that combines both inflammatory and thrombus biomarkers in the early phase of ischemic stroke (IS).

Methods

The study population consisted of 897 patients with a first diagnosis of IS admitted to the emergency department of five tertiary hospitals in China. Of these, the data from 70% of patients was randomly selected to derive the model and the other 30% for model validation. A TIPS of “2” was indicative of high inflammation and thrombosis biomarkers and “1” of one biomarker, with “0” indicative of absence of biomarkers. Multivariate logistic regression analyses were used to identify the association between TIPS and SAP.

Results

The TIPS was an independent predictor of SAP and 90-day mortality, with the incidence of SAP being significantly higher for patients with a high TIPS. The TIPS provided superior predictive value for SAP than clinical scores (A2DS2) and biomarkers currently used in practice, for both the derivation and validation sets. Mediation analysis revealed that TIPS provided a predictive value than either thrombotic (NLR) and inflammatory (D-dimer) biomarkers alone.

Conclusion

The TIPS score may be a useful tool for early identification of patients at high-risk for SAP after IS.

The Role of Arginine-Vasopressin in Stroke and the Potential Use of Arginine-Vasopressin Type 1 Receptor Antagonists in Stroke Therapy: A Narrative Review

Stroke is a life-threatening condition in which accurate diagnoses and timely treatment are critical for successful neurological recovery. The current acute treatment strategies, particularly non-invasive interventions, are limited, thus urging the need for novel therapeutical targets. Arginine vasopressin (AVP) receptor antagonists are emerging as potential targets to treat edema formation and subsequent elevation in intracranial pressure, both significant causes of mortality in acute stroke. Here, we summarize the current knowledge on the mechanisms leading to AVP hyperexcretion in acute stroke and the subsequent secondary neuropathological responses. Furthermore, we discuss the work supporting the predictive value of measuring copeptin, a surrogate marker of AVP in stroke patients, followed by a review of the experimental evidence suggesting AVP receptor antagonists in stroke therapy. As we highlight throughout the narrative, critical gaps in the literature exist and indicate the need for further research to understand better AVP mechanisms in stroke. Likewise, there are advantages and limitations in using copeptin as a prognostic tool, and the translation of findings from experimental animal models to clinical settings has its challenges. Still, monitoring AVP levels and using AVP receptor antagonists as an add-on therapeutic intervention are potential promises in clinical applications to alleviate stroke neurological consequences.

Neutrophil count multiplied by D-dimer combined with pneumonia may better predict short-term outcomes in patients with acute ischemic stroke

Objective

To investigate the predictive value of neutrophil, D-dimer and diseases associated with stroke for short-term outcomes of acute ischemic stroke (AIS).

Methods

By collecting the subitems of laboratory data especially routine blood and coagulation test in AIS patients, and recording their clinical status, the correlation, regression and predictive value of each subitem with the short-term outcomes of AIS were analyzed. The predict model was constructed.

Results

The neutrophil count multiplied by D-dimer (NDM) had the best predictive value among the subitems, and the area under the receiver operating characteristic (ROC) curve reached 0.804. When clinical information was not considered, the Youden index of NDM was calculated to be 0.48, corresponding to an NDM value of 7.78, a diagnostic sensitivity of 0.79, specificity of 0.69, negative predictive value of 96%. NDM were divided into 5 quintiles, the five grade of NDM (quintile) were < = 1.82, 1.83–2.41, 2.42–3.27, 3.28–4.49, 4.95+, respectively. The multivariate regression analysis was conducted between NDM (quintile), Babinski+, pneumonia, cardiac disease and poor outcomes of AIS. Compared with the first grade of NDM (quintile), the second grade of NDM (quintile) was not significant, but the third grade of NDM (quintile) showed 7.061 times, the fourth grade of NDM (quintile) showed 11.776 times, the fifth grade of NDM (quintile) showed 23.394 times in short-term poor outcomes occurrence. Babinski sign + showed 1.512 times, pneumonia showed 2.995 times, cardiac disease showed 1.936 times in short-term poor outcomes occurrence compared with those negative patients.

Conclusions

NDM combined with pneumonia may better predict short-term outcomes in patients with AIS. Early prevention, regular examination and timely intervention should be emphasized for patients, which may reduce the risk of short-term poor outcomes.

Novel machine learning models to predict pneumonia events in supratentorial intracerebral hemorrhage populations: An analysis of the Risa-MIS-ICH study

Background

Stroke-associated pneumonia (SAP) contributes to high mortality rates in spontaneous intracerebral hemorrhage (sICH) populations. Accurate prediction and early intervention of SAP are associated with prognosis. None of the previously developed predictive scoring systems are widely accepted. We aimed to derive and validate novel supervised machine learning (ML) models to predict SAP events in supratentorial sICH populations.

Methods

The data of eligible supratentorial sICH individuals were extracted from the Risa-MIS-ICH database and split into training, internal validation, and external validation datasets. The primary outcome was SAP during hospitalization. Univariate and multivariate analyses were used for variable filtering, and logistic regression (LR), Gaussian naïve Bayes (GNB), random forest (RF), K-nearest neighbor (KNN), support vector machine (SVM), extreme gradient boosting (XGB), and ensemble soft voting model (ESVM) were adopted for ML model derivations. The accuracy, sensitivity, specificity, and area under the curve (AUC) were adopted to evaluate the predictive value of each model with internal/cross-/external validations.

Results

A total of 468 individuals with sICH were included in this work. Six independent variables [nasogastric feeding, airway support, unconscious onset, surgery for external ventricular drainage (EVD), larger sICH volume, and intensive care unit (ICU) stay] for SAP were identified and selected for ML prediction model derivations and validations. The internal and cross-validations revealed the superior and robust performance of the GNB model with the highest AUC value (0.861, 95% CI: 0.793–0.930), while the LR model had the highest AUC value (0.867, 95% CI: 0.812–0.923) in external validation. The ESVM method combining the other six methods had moderate but robust abilities in both cross-validation and external validation and achieved an AUC of 0.843 (95% CI: 0.784–0.902) in external validation.

Conclusion

The ML models could effectively predict SAP in sICH populations, and our novel ensemble model demonstrated reliable robust performance outcomes despite the populational and algorithmic differences. This attempt indicated that ML application may benefit in the early identification of SAP.

Hypersensitive C-reactive protein-albumin ratio is associated with stroke-associated pneumonia and early clinical outcomes in patients with acute ischemic stroke

OBJECTIVES

We aimed to explore the association between the baseline hypersensitive C-reactive protein-albumin ratio (CAR) and stroke-associated pneumonia (SAP) during hospitalization and the short-term prognosis in patients with acute ischemic stroke (AIS).

METHODS

We enrolled 766 patients with AIS and collected their admission baseline characteristics, including their National Institutes of Health Stroke Scale score, CAR, age, atrial fibrillation, dysphagia, sex, stroke severity (A² DS² ) score, and other information. The occurrence of SAP within 7 days after stroke, length of hospital stay, and physical condition at discharge were also recorded. The patients' Modified Rankin Scale (mRS) scores and mortality 3 months after AIS were further evaluated at follow-up. All patients were divided into four groups based on the quartiles of the admission CAR (Q1 <1.3, Q2 1.3-3.7, Q3 3.7-9.3, Q4 ≥9.3).

RESULTS

During hospitalization, 92 (11.9%) patients were diagnosed with SAP. The patients with SAP had a higher CAR than the non-SAP patients (p < .001). In the multivariate-adjusted model, the patients in the Q3 and Q4 groups had a higher SAP risk (aOR was 5.21 and 17.72, p-trend < .001) than those in the lowest quartile. The area under the curve for the CAR's ability to predict SAP was 0.810 in the receiver operating characteristic curve analysis and had a similar predictive efficacy as the A² DS²  score (p <.05). The length of stay in the SAP group was almost the same as that in the non-SAP group, but the clinical outcomes were worse at discharge and at the 3-month follow-up in the SAP group. In addition, the patients in the higher CAR quartiles at admission were more likely to have poorer clinical outcomes.

CONCLUSIONS

Patients with AIS with a high CAR at admission are more likely to develop SAP during hospitalization and have poor short-term clinical outcomes. These findings might help to timely identify patients at high risk of SAP and provide a basis for further research on prophylactic antibiotic therapy.

ICH-LR2S2: a new risk score for predicting stroke-associated pneumonia from spontaneous intracerebral hemorrhage

Purpose

We develop a new risk score to predict patients with stroke-associated pneumonia (SAP) who have an acute intracranial hemorrhage (ICH).

Method

We applied logistic regression to develop a new risk score called ICH-LR2S2. It was derived from examining a dataset of 70,540 ICH patients between 2015 and 2018 from the Chinese Stroke Center Alliance (CSCA). During the training of ICH-LR2S2, patients were randomly divided into two groups – 80% for the training set and 20% for model validation. A prospective test set was developed using 12,523 patients recruited in 2019. To further verify its effectiveness, we tested ICH-LR2S2 on an external dataset of 24,860 patients from the China National Stroke Registration Management System II (CNSR II). The performance of ICH-LR2S2 was measured by the area under the receiver operating characteristic curve (AUROC).

Results

The incidence of SAP in the dataset was 25.52%. A 24-point ICH-LR2S2 was developed from independent predictors, including age, modified Rankin Scale, fasting blood glucose, National Institutes of Health Stroke Scale admission score, Glasgow Coma Scale score, C-reactive protein, dysphagia, Chronic Obstructive Pulmonary Disease, and current smoking. The results showed that ICH-LR2S2 achieved an AUC = 0.749 [95% CI 0.739–0.759], which outperforms the best baseline ICH-APS (AUC = 0.704) [95% CI 0.694–0.714]. Compared with the previous ICH risk scores, ICH-LR2S2 incorporates fasting blood glucose and C-reactive protein, improving its discriminative ability. Machine learning methods such as XGboost (AUC = 0.772) [95% CI 0.762–0.782] can further improve our prediction performance. It also performed well when further validated by the external independent cohort of patients (n = 24,860), ICH-LR2S2 AUC = 0.784 [95% CI 0.774–0.794].

Conclusion

ICH-LR2S2 accurately distinguishes SAP patients based on easily available clinical features. It can help identify high-risk patients in the early stages of diseases.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-022-03389-5.

Novel Predictors of Stroke-Associated Pneumonia: A Single Center Analysis

Stroke-associated pneumonia (SAP) is a common cause of disability or death. Although the researches on SAP have been relatively mature, the method that can predict SAP with great accuracy has not yet been determined. It is necessary to discover new predictors to construct a more accurate predictive model for SAP. We continuously collected 2,366 patients with acute ischemic stroke, and then divided them into the SAP group and non-SAP group. Data were recorded at admission. Through univariate analyses and multivariate regression analyses of the data, the new predictive factors and the predictive model of SAP were determined. The receiver operating characteristic (ROC) curve and the corresponding area under the curve (AUC) were used to measure their predictive accuracy. Of the 2,366 patients, 459 were diagnosed with SAP. International normalized ratio (INR) (odds ratio = 37.981; 95% confidence interval, 7.487–192.665; P &lt; 0.001), age and dysphagia were independent risk factors of SAP. However, walking ability within 48 h of admission (WA) (odds ratio = 0.395; 95% confidence interval, 0.287–0.543; P &lt; 0.001) was a protective factor of SAP. Different predictors and the predictive model all could predict SAP (P &lt; 0.001). The predictive power of the model (AUC: 0.851) which included age, homocysteine, INR, history of chronic obstructive pulmonary disease (COPD), dysphagia, and WA was greater than that of age (AUC: 0.738) and INR (AUC: 0.685). Finally, we found that a higher INR and no WA could predict SAP in patients with acute ischemic stroke. In addition, we designed a simple and practical predictive model for SAP, which showed relatively good accuracy. These findings might help identify high-risk patients with SAP and provide a reference for the timely use of preventive antibiotics.

Immunodepression, Infections, and Functional Outcome in Ischemic Stroke

Stroke remains one of the main causes of mortality and morbidity worldwide. Immediately after stroke, a neuroinflammatory process starts in the brain, triggering a systemic immunodepression mainly through excessive activation of the autonomous nervous system. Manifestations of immunodepression include lymphopenia but also dysfunctional innate and adaptive immune cells. The resulting impaired antibacterial defenses render patients with stroke susceptible to infections. In addition, other risk factors like stroke severity, dysphagia, impaired consciousness, mechanical ventilation, catheterization, and older age predispose stroke patients for infections. Most common infections are pneumonia and urinary tract infection, both occur in ≈10% of the patients. Especially pneumonia increases unfavorable outcome and mortality in patients with stroke; systemic effects like hypotension, fever, delay in rehabilitation are thought to play a crucial role. Experimental and clinical data suggest that systemic infections enhance autoreactive immune responses against brain antigens and thus negatively affect outcome but convincing evidence is lacking. Prevention of poststroke infections by preventive antibiotic therapy did not improve functional outcome after stroke. Immunomodulatory approaches counteracting immunodepression to prevent stroke-associated pneumonia need to account for neuroinflammation in the ischemic brain and avoid further tissue damage. Experimental studies discovered interesting targets, but these have not yet been investigated in patients with stroke. A better understanding of the pathobiology may help to develop optimized approaches of preventive antibiotic therapy or immunomodulation to effectively prevent stroke-associated pneumonia while improving long-term outcome after stroke. In this review, we aim to characterize epidemiology, risk factors, cause, diagnosis, clinical presentation, and potential treatment of poststroke immunosuppression and associated infections.

A Combined Clinical and Serum Biomarker-Based Approach May Allow Early Differentiation Between Patients With Minor Stroke and Transient Ischemic Attack as Well as Mid-term Prognostication

Background: Early differentiation between transient ischemic attack (TIA) and minor ischemic stroke (MIS) impacts on the patient's individual diagnostic work-up and treatment. Furthermore, estimations regarding persisting impairments after MIS are essential to guide rehabilitation programs. This study evaluated a combined clinical- and serum biomarker-based approach for the differentiation between TIA and MIS as well as the mid-term prognostication of the functional outcome, which is applicable within the first 24 h after symptom onset. Methods: Prospectively collected data were used for a retrospective analysis including the neurological deficit at admission (National Institutes of Health Stroke Scale, NIHSS) and the following serum biomarkers covering different pathophysiological aspects of stroke: Coagulation (fibrinogen, antithrombin), inflammation (C reactive protein), neuronal damage in the cellular [neuron specific enolase], and the extracellular compartment [matrix metalloproteinase-9, hyaluronic acid]. Further, cerebral magnetic resonance imaging was performed at baseline and day 7, while functional outcome was evaluated with the modified Rankin Scale (mRS) after 3, 6, and 12 months. Results: Based on data from 96 patients (age 64 ± 14 years), 23 TIA patients (NIHSS 0.6 ± 1.1) were compared with 73 MIS patients (NIHSS 2.4 ± 2.0). In a binary logistic regression analysis, the combination of NIHSS and serum biomarkers differentiated MIS from TIA with a sensitivity of 91.8% and a specificity of 60.9% [area under the curve (AUC) 0.84]. In patients with NIHSS 0 at admission, this panel resulted in a still acceptable sensitivity of 81.3% (specificity 71.4%, AUC 0.69) for the differentiation between MIS (n = 16) and TIA (n = 14). By adding age, remarkable sensitivities of 98.4, 100, and 98.2% for the prediction of an excellent outcome (mRS 0 or 1) were achieved with respect to time points investigated within the 1-year follow-up. However, the specificity was moderate and decreased over time (83.3, 70, 58.3%; AUC 0.96, 0.92, 0.91). Conclusion: This pilot study provides evidence that the NIHSS combined with selected serum biomarkers covering pathophysiological aspects of stroke may represent a useful tool to differentiate between MIS and TIA within 24 h after symptom onset. Further, this approach may accurately predict the mid-term outcome in minor stroke patients, which might help to allocate rehabilitative resources.

Breath and plasma metabolomics to assess inflammation in acute stroke

Inflammation is strongly implicated in both injury and repair processes occurring after stroke. In this exploratory study we assessed the feasibility of repeated sampling of exhaled volatile organic compounds and performed an untargeted metabolomic analysis of plasma collected at multiple time periods after stroke. Metabolic profiles were compared with the time course of the inflammatory markers C-reactive protein (CRP) and interleukin-6 (IL-6). Serial breath sampling was well-tolerated by all patients and the measurement appears feasible in this group. We found that exhaled decanal tracks CRP and IL-6 levels post-stroke and correlates with several metabolic pathways associated with a post-stroke inflammatory response. This suggests that measurement of breath and blood metabolites could facilitate development of novel therapeutic and diagnostic strategies. Results are discussed in relation to the utility of breath analysis in stroke care, such as in monitoring recovery and complications including stroke associated infection.

Copeptin and Stress

Vasopressin (AVP) and copeptin are released in equimolar amounts from the same precursor. Due to its molecular stability and countless advantages as compared with AVP, copeptin perfectly mirrors AVP presence and has progressively emerged as a reliable marker of vasopressinergic activation in response to osmotic and hemodynamic stimuli in clinical practice. Moreover, evidence highlighting the prognostic potential of copeptin in several acute diseases, where the activation of the AVP system is primarily linked to stress, as well as in psychologically stressful conditions, has progressively emerged. Furthermore, organic stressors induce a rise in copeptin levels which, although non-specific, is unrelated to plasma osmolality but proportional to their magnitude: suggesting disease severity, copeptin proved to be a reliable prognostic biomarker in acute conditions, such as sepsis, early post-surgical period, cardiovascular, cerebrovascular or pulmonary diseases, and even in critical settings. Evidence on this topic will be briefly discussed in this article.

The Relationship Between Serum YKL-40 Levels on Admission and Stroke-Associated Pneumonia in Patients with Acute Ischemic Stroke

Background

Stroke-associated pneumonia (SAP) is a standout complication after acute ischemic stroke (AIS), with a prevalence of 7–38%. The aim of this prospective study was to investigate the relationship between serum YKL-40 levels at admission and SAP.

Methods

Between August 2020 and February 2021, consecutive AIS patients from two centers were enrolled prospectively. Serum YKL-40 concentrations were measured via enzyme-linked immunosorbent assay. We performed logistic regression analyses to explore the relationship between YKL-40 and SAP. Receiver operating characteristic curve was also used to assess the predictive ability of YKL-40 in predicting SAP.

Results

Ultimately, a total of 511 AIS patients were recruited. Multivariate logistic regression analysis showed that YKL-40 was independently related to SAP, whether as a continuous variable or as quartiles (P=0.001). The area under curve of YKL-40 to predict SAP was 0.765. The optimal cutoff value of YKL-40 as a predictor of SAP was determined to be 206.4 ng/mL, where the sensitivity was 63.1% and the specificity was 82.0%.

Conclusion

Our study demonstrated that YKL-40 might be considered as a useful biomarker to predict SAP in AIS patients.

Lipopolysaccharide binding protein and sCD14 as risk markers of stroke-associated pneumonia

To determine the utility of lipopolysaccharide binding protein (LBP) and soluble CD14 (sCD14) as risk markers of stroke-associated pneumonia (SAP). We included 331 stroke patients. The plasma levels of LBP (median: 19.4 vs 15.3 μg/mL, P < 0.01) and sCD14 (median: 1.5 vs 1.4 μg/mL, P = 0.04) were elevated in SAP. In multivariate analysis, a higher level of LBP (OR: 1.09, 95%CI: 1.05-1.13), but not sCD14 (OR: 2.16, 0.94-4.97), was associated with SAP. The addition of LBP or sCD14 to the clinical model did not improve its discriminatory ability. Our results suggest the modest value of studied biomarkers for SAP prediction.