Acta Cardiologica Sinica最新文献

Background: Venoarterial (V-A) extracorporeal membrane oxygenation (ECMO) after cardiac arrest often predisposes patients to acute brain injury (ABI), which affects survival and neurological performance. The investigation of the predictors of ABI will be beneficial for further management.

Objectives: To explore the predictors and outcomes of ABI and intracerebral hemorrhage (ICH) in patients experiencing cardiac arrest and cardiopulmonary resuscitation (CPR) with V-A ECMO support.

Methods: We retrospectively analyzed 150 patients who successfully weaned from V-A ECMO support after pre-ECMO CPR at our institution from January 2009 to December 2021. Short-term and long-term outcomes were evaluated. Characteristics before and during ECMO were analyzed for determining the predictors of ABI and ICH.

Results: Of the 150 patients, 66 (44.0%) had ABI. ABI was associated with higher in-hospital mortality (62.1% vs. 21.4%, p < 0.0001) and poorer long-term survival after discharge (p = 0.002). Patients who survived to discharge with ABI had significantly more severe neurological deficits at discharge (84.0% vs. 42.4%, p < 0.0001) and improved little at one year after discharge (33.3% vs. 11.4%, p = 0.027). We found that CPR duration [odds ratio (OR) = 1.04, p = 0.003] was the independent risk factor for ABI, whereas lower platelet counts was the independent risk factor for ICH (OR = 0.96, p = 0.019).

Conclusions: After CPR, development of ABI during V-A ECMO support impacted survival and further neurological outcome. Longer CPR duration before ECMO set up significantly increases the occurrence of ABI. Besides, severe thrombocytopenia during ECMO support increases the possibility of ICH.

Background: The efficacy of the left atrial (LA) expansion index (LAEI) to predict cerebral ischemic events in patients with atrial fibrillation (AF) is unknown.

Methods: We enrolled 177 patients with AF (88 with paroxysmal AF and 89 with persistent AF) and a baseline CHA₂DS₂-VASc score (at enrollment) of 3.6 ± 2.3. Comprehensive echocardiography was performed at enrollment. The LAEI was calculated as (Vol_max - Vol_min) × 100%/Vol_min, where Vol_max and Vol_min denoted maximal and minimal LA volumes, respectively. The study endpoint was ischemic stroke. Stroke subtypes were classified into cardioembolic stroke (CE), non-CE with determined mechanism (NCE), embolic stroke of undetermined source (ESUS), or transient ischemic attack (TIA).

Results: Over a mean 9.9-year follow-up period, 44 (24.9%) of the patients reached the endpoint (24 with CE, 4 with NCE, 6 with ESUS, and 10 with TIA). The LAEI was lower in the stroke group than in the non-stroke group. Stroke incidence in the lowest LAEI quartile was much higher than that in the other LAEI quartiles; the 10-year cumulative stroke risk was 15.9% (14/88) and 33.7% (30/89) in the patients with paroxysmal and persistent AF, respectively. An LAEI of < 35% predicted the presence of stroke with 77% sensitivity and 78% specificity. In multivariable analysis, the LAEI was independently associated with ischemic stroke (hazard ratio 0.952 per 1% increase, 95% confidence interval 0.932-0.971, p < 0.0001).

Conclusions: The LAEI is a useful predictor of ischemic stroke in patients with AF.

Background: Serum albumin (SA), a multifunction protein, contributes to maintaining a variety of physiological functions. Studies have linked SA to atherosclerosis with possible mechanisms including a response to inflammation. The contribution of albumin to cardiovascular (CV) mortality in elderly patients with stable coronary artery disease (CAD) remains unclear.

Methods: We investigated 321 elderly patients with stable CAD undergoing coronary angiography between 2003 and 2006. CV mortality data were obtained from the National Registry of Deaths in Taiwan. CV mortality included deaths attributable to ischemic heart disease, congestive heart disease, and stroke. The association between baseline SA and CV mortality was assessed using a Cox model and Fine-Gray model when non-CV mortality was considered a competing event.

Results: During a median follow-up of 97 months, 39 (12.1%) participants died from CV disease and 76 (23.7%) died from non-CV diseases. After adjusting for covariates, patients in the SA ≥ 3.75 g/dL group had a lower frequency of CV mortality compared with those in the SA < 3.75 g/dL group [hazard ratio (HR): 0.20; 95% confidence interval (CI): 0.08-0.49; p < 0.001]. Similarly, compared to the participants with non-CV mortality, the SA ≥ 3.75 g/dL group had a lower frequency of CV mortality compared with the SA < 3.75 g/dL group (subdistribution HR: 0.27; 95% CI: 0.11-0.65; p < 0.001) in adjusted competing risk models.

Conclusions: A SA level ≥ 3.75 g/dL at admission was associated with decreased long-term CV mortality and may be useful for risk prediction in elderly patients with stable CAD.

Background: Left atrial strain can usefully reflect left atrial function. The follow-up periods in previous studies assessing left atrial strain as a survival predictor have been relatively short, and few studies have examined the ability of left atrial strain to predict mortality in patients with borderline diastolic function. This study sought to investigate the survival predictive value of left atrial strain with a longer follow-up duration. In addition, we also evaluated the survival predictive value of left atrial strain in patients with borderline diastolic function.

Methods: In total, 652 participants who received routine echocardiography underwent 2-D speckle tracking echocardiography to evaluate left atrial reservoir function by peak atrial longitudinal strain. The study endpoints were all-cause and cardiovascular mortality.

Results: The mean left atrial strain was 27.6%, and the median follow-up duration was 92 months. During follow-up, 72 patients died of cardiovascular causes and 181 died of all causes. Univariable Cox regression analysis revealed that lower left atrial strain significantly predicted an increase in all-cause and cardiovascular mortality. After adjusting for common clinical and echocardiographic parameters, lower left atrial strain was still associated with a higher risk of all-cause mortality [hazard ratio (HR) = 0.942, p = 0.011] and cardiovascular mortality (HR = 0.915, p = 0.018) in multivariable Cox-regression analysis. In addition, 293 patients had borderline left ventricular diastolic function. Multivariable analysis still revealed that left atrial strain could predict cardiovascular mortality in this population.

Conclusions: Our data showed that left atrial strain could predict all-cause and cardiovascular mortality, even after adjusting for general clinical and echocardiographic parameters.

Cardiac rehabilitation is a comprehensive intervention recommended in international and Taiwanese guidelines for patients with acute myocardial infarction. Evidence supports that cardiac rehabilitation improves the health-related quality of life, enhances exercise capacity, reduces readmission rates, and promotes survival in patients with cardiovascular disease. The cardiac rehabilitation team is comprehensive and multidisciplinary. The inpatient, outpatient, and maintenance phases are included in cardiac rehabilitation. All patients admitted with acute myocardial infarction should be referred to the rehabilitation department as soon as clinically feasible. Pre-exercise evaluation, including exercise testing, helps physicians identify the risks of cardiac rehabilitation and organize appropriate exercise prescriptions. Therefore, the Taiwan Myocardial Infarction Society (TAMIS), Taiwan Society of Cardiology (TSOC), and Taiwan Academy of Cardiovascular and Pulmonary Rehabilitation (TACVPR) address this consensus statement to assist healthcare practitioners in performing cardiac rehabilitation in patients with acute myocardial infarction.

Background: The early diagnosis of pulmonary embolism (PE) remains a challenge. Electrocardiograms (ECGs) and D-dimer levels are used to screen potential cases.

Objective: To develop a deep learning model (DLM) to detect PE using ECGs and investigate the clinical value of false detections in patients without PE.

Methods: Among patients who visited the emergency department between 2011 and 2019, PE cases were identified through a review of medical records. Non-PE ECGs were collected from patients without a diagnostic code for PE. There were 113 PE and 51,456 non-PE ECGs in the training and validation sets for developing the DLM, respectively, and 27 PE and 13,105 non-PE cases in an independent testing set for performance validation. A human-machine competition was conducted from the testing set to compare the performance of the DLM with that of physicians. Receiver operating characteristic (ROC) curves, sensitivity, and specificity were used to determine the diagnostic value. Survival analysis was used to assess the prognosis of the patients without PE, stratified by DLM prediction.

Results: The DLM was as effective as physicians in diagnosing PE, with 70.8% sensitivity and 69.7% specificity. The area under the ROC curve of DLM was 0.778 in the testing set and up to 0.9 with D-dimer and demographic data. The non-PE patients whose ECG was misclassified as PE by DLM had higher all-cause mortality [hazard ratio (HR) 2.13 (1.51-3.02)] and risk of non-cardiovascular hospitalization [HR 1.55 (1.42-1.68)] than those correctly classified.

Conclusions: A DLM-enhanced ECG system may prompt PE recognition and provide prognostic outcomes in patients with false-positive predictions.

Background: The 10-year atherosclerotic cardiovascular disease (ASCVD) risk - as assessed using the Framingham general cardiovascular risk score (FRS-CVD) or pooled cohort equations (PCE) - is commonly used in Western cohorts for the primary prevention of cardiovascular disease (CVD). However, the FRS-CVD and PCE have not been validated in Taiwanese cohorts.

Objectives: We aimed to validate the FRS-CVD and PCE for assessing the 10-year ASCVD risk using a Taiwanese community-based population.

Methods: We extracted patient data from the Landseed Integrated Outreaching Neighborhood Screening registry, a community-based prospective cohort study established in 2006. Cardiovascular events from 2006 to 2017 were determined from electronic medical records. The discriminative power and calibration of the FRS-CVD and PCE were evaluated.

Results: Overall, 5,139 subjects were analyzed; the 10-year follow-up rate was 99.6%. The mean age at baseline was 52.8 ± 13.1 years, and 44.6% of the subjects were male. In total, 430 of 4,631 (9.3%) and 227 of 4,022 (5.6%) of the FRS-CVD- and PCE-like cohorts, respectively, had ASCVD events. The calibration χ² of the FRS-CVD was 7.0267 (p = 0.6343) in males and 7.8845 (p = 0.5458) in females; the χ² of PCE was 13.007 (p = 0.1623) in males and 38.785 (p < 0.001) in females. The area under the receiver operating characteristic curve (AUROC) of the FRS-CVD was 0.76 (0.72-0.79) in males and 0.71 (0.67-0.74) in females; the AUROC of PCE was 0.68 (0.62-0.73) in males and 0.61 (0.56-0.67) in females.

Conclusions: Except for PCE in females, the FRS-CVD and PCE provided good calibration and modest discrimination in statin-naïve Taiwanese individuals without prior CVD.

Background: Galectin-3 affects cardiac tissue inflammation as an inflammatory mediator. The development of cardiorenal syndrome in heart failure patients is associated with a poor prognosis. This study aims to investigate whether serum galectin-3 levels can be used as a biomarker to predict cardiorenal syndrome in heart failure patients with reduced left ventricular ejection fraction.

Methods: A total of 166 symptomatic heart failure patients [New York Heart Association (NYHA) functional class II-III] with reduced left ventricular ejection fraction (≤ 40%) were recruited prospectively. Cardiorenal syndrome type 1 was defined as an acute worsening of cardiac function leading to renal dysfunction. The patients were divided into two groups with and without cardiorenal syndrome. The galectin-3 levels of all patients were determined. The primary outcome of this study was the occurrence of cardiorenal syndrome.

Results: Cardiorenal syndrome developed in 41 patients. Galectin-3 levels were found to be higher in the patients with cardiorenal syndrome (+) compared to those without cardiorenal syndrome (-) (20.7 ± 2.9 ng/mL vs. 17.8 ± 3.1 ng/mL, p < 0.001). After performing a multivariable analysis, galectin-3 levels [odds ratio (OR): 3.21, p = 0.001], NYHA functional class (OR: 1.98, p = 0.009), creatinine (OR: 3.18, p = 0.006), furosemide dose (OR: 1.21, p = 0.033), and angiotensin-converting enzyme inhibitor/angiotensin-receptor blockers usage (OR: 0.54, p = 0.029) were identified as independent predictors for the development of cardiorenal syndrome. Moreover, galectin-3 level demonstrated predictive capability for cardiorenal syndrome development (AUC = 0.761, p < 0.001).

Conclusions: Serum galectin-3 level showed an association with cardiorenal syndrome development in patients with heart failure, indicating potential usefulness as a prognostic biomarker.

Introduction: Atherosclerotic cardiovascular disease (ASCVD) is prevalent worldwide including Taiwan, however widely accepted tools to assess the risk of ASCVD are lacking in Taiwan. Machine learning models are potentially useful for risk evaluation. In this study we used two cohorts to test the feasibility of machine learning with transfer learning for developing an ASCVD risk prediction model in Taiwan.

Methods: Two multi-center observational registry cohorts, T-SPARCLE and T-PPARCLE were used in this study. The variables selected were based on European, U.S. and Asian guidelines. Both registries recorded the ASCVD outcomes of the patients. Ten-fold validation and temporal validation methods were used to evaluate the performance of the binary classification analysis [prediction of major adverse cardiovascular (CV) events in one year]. Time-to-event analyses were also performed.

Results: In the binary classification analysis, eXtreme Gradient Boosting (XGBoost) and random forest had the best performance, with areas under the receiver operating characteristic curve (AUC-ROC) of 0.72 (0.68-0.76) and 0.73 (0.69-0.77), respectively, although it was not significantly better than other models. Temporal validation was also performed, and the data showed significant differences in the distribution of various features and event rate. The AUC-ROC of XGBoost dropped to 0.66 (0.59-0.73), while that of random forest dropped to 0.69 (0.62-0.76) in the temporal validation method, and the performance also became numerically worse than that of the logistic regression model. In the time-to-event analysis, most models had a concordance index of around 0.70.

Conclusions: Machine learning models with appropriate transfer learning may be a useful tool for the development of CV risk prediction models and may help improve patient care in the future.