A statistical predictive model consistent within a 5-year follow-up period for patients with acute heart failure

Using ultrasound radiomics to forecast adverse cardiovascular events in patients with acute coronary syndrome after percutaneous coronary intervention

OBJECTIVE

Exploring the performance of ultrasound-based radiomics in forecasting major adverse cardiovascular events (MACE) within 1 year following percutaneous coronary intervention (PCI) of acute coronary syndrome (ACS) patients.

METHODS

In this research, 161 ACS patients who underwent PCI were included (114 patients were randomly assigned to the training set and 47 patients to the validation set). Every patient received echocardiography 3-7 days after PCI and followed up for 1 year. The radiomics features related to MACE occurrence were extracted and selected to formulate the RAD score. Building ultrasound personalized model by incorporating RAD score, LVEF, LVGLS, and NT-ProBNP. The model's capacity to predict was tested using ROC curves.

RESULTS

Multifactorial logistic regression analysis of RAD score with clinical data and echocardiographic parameters indicated RAD score and LVGLS as independent risk factors for the occurrence of MACE. The RAD score predicted MACE, with AUC values of 0.85 and 0.86 in the training and validation sets. The ultrasound personalized model had a superior ability to predict the occurrence of MACE, with AUC values of 0.88 and 0.92, which were higher than those of the clinical model (with AUC of 0.72 and 0.80) without RAD score (Z = 3.711, 2.043, P < .001, P = .041). Furthermore, DCA indicated that the ultrasound personalization model presented a more favorable net clinical benefit.

CONCLUSIONS

Ultrasound radiomics can be a reliable tool to predict the incidence of MACE after PCI in patients with ACS and provides quantifiable data for personalized clinical treatment.

Prognostic impact of MitraScore in elderly Asian patients with heart failure: sub-analysis of FRAGILE-HF

AIMS

MitraScore is a novel, simple, and manually calculatable risk score developed as a prognostic model for patients undergoing transcatheter edge-to-edge repair (TEER) for mitral regurgitation. As its components are considered prognostic in heart failure (HF), we aimed to investigate the usefulness of the MitraScore in HF patients.

METHODS AND RESULTS

We calculated MitraScore for 1100 elderly patients (>65 years old) hospitalized for HF in the prospective multicentre FRAGILE-HF study and compared its prognostic ability with other simple risk scores. The primary endpoint was all-cause deaths, and the secondary endpoints were the composite of all-cause deaths and HF rehospitalization and cardiovascular deaths. Overall, the mean age of 1100 patients was 80 ± 8 years, and 58% were men. The mean MitraScore was 3.2 ± 1.4, with a median of 3 (interquartile range: 2-4). A total of 326 (29.6%), 571 (51.9%), and 203 (18.5%) patients were classified into low-, moderate-, and high-risk groups based on the MitraScore, respectively. During a follow-up of 2 years, 226 all-cause deaths, 478 composite endpoints, and 183 cardiovascular deaths were observed. MitraScore successfully stratified patients for all endpoints in the Kaplan-Meier analysis (P < 0.001 for all). In multivariate analyses, MitraScore was significantly associated with all endpoints after covariate adjustments [adjusted hazard ratio (HR) (95% confidence interval): 1.22 (1.10-1.36), P < 0.001 for all-cause deaths; adjusted HR 1.17 (1.09-1.26), P < 0.001 for combined endpoints; and adjusted HR 1.24 (1.10-1.39), P < 0.001 for cardiovascular deaths]. The Hosmer-Lemeshow plot showed good calibration for all endpoints. The net reclassification improvement (NRI) analyses revealed that the MitraScore performed significantly better than other manually calculatable risk scores of HF: the GWTG-HF risk score, the BIOSTAT compact model, the AHEAD score, the AHEAD-U score, and the HANBAH score for all-cause and cardiovascular deaths, with respective continuous NRIs of 0.20, 0.22, 0.39, 0.39, and 0.29 for all-cause mortality (all P-values < 0.01) and 0.20, 0.22, 0.42, 0.40, and 0.29 for cardiovascular mortality (all P-values < 0.02).

CONCLUSIONS

MitraScore developed for patients undergoing TEER also showed strong discriminative power in HF patients. MitraScore was superior to other manually calculable simple risk scores and might be a good choice for risk assessment in clinical practice for patients receiving TEER and those with HF.

Usefulness of HANBAH Score in Japanese Patients With Acute Heart Failure

The HANBAH score is a novel simple risk score consisting of hemoglobin level, age, sodium (N) level, blood urea nitrogen level, atrial fibrillation, and high-density lipoprotein. We aimed to validate this score in an external population. This retrospective study included 744 patients hospitalized for acute heart failure between 2015 and 2019. Each of the following criteria was scored as 1 point: hemoglobin level (<13.0 g/L for men and <12.0 g/L for women), atrial fibrillation, age (>70 years), serum blood urea nitrogen level (>26 mg/100 ml for men and >28 mg/100 ml for women), serum high-density lipoprotein level (<25 mg/100 ml), and serum sodium level (<135 mg/100 ml). HANBAH scores were available for 736 patients (age, 75 ± 13 years; 60% male; reduced [<40%] and preserved ejection fraction [≥50%]: 35% and 49%, respectively). All-cause death during follow-up, a composite of death and heart failure rehospitalization, and in-hospital death were observed in 173, 274, and 51 patients, respectively. The HANBAH score was significantly associated with these end points after adjustment for covariates (adjusted hazard ratio 1.38 [95% confidence interval 1.16 to 1.64], p <0.001; 1.27 [1.11 to 1.45], p <0.001; and 1.66 [1.18 to 2.33], p <0.001, respectively). Receiver operating characteristic and net reclassification improvement analyses showed that the HANBAH score performed significantly better than AHEAD (atrial fibrillation, hemoglobin [anemia], elderly, abnormal renal parameters, diabetes mellitus) and AHEAD-U (AHEAD with uric acid) scores and similar to the multi-domain ACUTE HF score for all end points. In conclusion, the HANBAH score showed powerful risk stratification in this external Japanese cohort. Despite its simplicity, it performed better than other simple risk scores and similar to a multidomain risk score.

Virtual healthcare solutions in heart failure: a literature review

The widespread adoption of mobile technologies offers an opportunity for a new approach to post-discharge care for patients with heart failure (HF). By enabling non-invasive remote monitoring and two-way, real-time communication between the clinic and home-based patients, as well as a host of other capabilities, mobile technologies have a potential to significantly improve remote patient care. This literature review summarizes clinical evidence related to virtual healthcare (VHC), defined as a care team + connected devices + a digital solution in post-release care of patients with HF. Searches were conducted on Embase (06/12/2020). A total of 171 studies were included for data extraction and evidence synthesis: 96 studies related to VHC efficacy, and 75 studies related to AI in HF. In addition, 15 publications were included from the search on studies scaling up VHC solutions in HF within the real-world setting. The most successful VHC interventions, as measured by the number of reported significant results, were those targeting reduction in rehospitalization rates. In terms of relative success rate, the two most effective interventions targeted patient self-care and all-cause hospital visits in their primary endpoint. Among the three categories of VHC identified in this review (telemonitoring, remote patient management, and patient self-empowerment) the integrated approach in remote patient management solutions performs the best in decreasing HF patients' re-admission rates and overall hospital visits. Given the increased amount of data generated by VHC technologies, artificial intelligence (AI) is being investigated as a tool to aid decision making in the context of primary diagnostics, identifying disease phenotypes, and predicting treatment outcomes. Currently, most AI algorithms are developed using data gathered in clinic and only a few studies deploy AI in the context of VHC. Most successes have been reported in predicting HF outcomes. Since the field of VHC in HF is relatively new and still in flux, this is not a typical systematic review capturing all published studies within this domain. Although the standard methodology for this type of reviews was followed, the nature of this review is qualitative. The main objective was to summarize the most promising results and identify potential research directions.

External validation of the ACUTE HF score for risk stratification in acute heart failure

BACKGROUND

The ACUTE HF score is a simple risk score that predicts the prognosis of patients with acute heart failure (HF) using clinical and echocardiographic parameters. As this score was developed for a small European population, we aimed to validate this score in an external population.

METHODS AND RESULTS

This retrospective observational cohort analysis included patients hospitalized with acute HF during 2015-2019. Of 744 patients, 703 patients with available ACUTE HF scores were analyzed (75 ± 13 years; 61% male; left ventricular ejection fraction [LVEF] 49 ± 17%). Approximately one-third (34.4%) of the patients had reduced LVEF (<40%), and 51.4% exhibited preserved LVEF (≥50%). During a median follow-up of 452 days, primary and secondary outcomes were observed in 110 and 204 patients, respectively. The ACUTE HF score successfully stratified patients for primary (all-cause mortality) and secondary endpoints (a composite of all-cause mortality and heart failure rehospitalization) in Kaplan-Meier analyses (log-rank test, P < 0.001). Multivariable Cox proportional hazards models showed that the score was significantly independently associated with both primary and secondary endpoints after adjusted by covariates (P < 0.001).

CONCLUSION

We validated the risk prediction ability of ACUTE HF score in an Asian population. This score may be applicable in clinical practice.

Predictive Model and Mortality Risk Score during Admission for Ischaemic Stroke with Conservative Treatment

Background: Stroke is the second cause of mortality worldwide and the first in women. The aim of this study is to develop a predictive model to estimate the risk of mortality in the admission of patients who have not received reperfusion treatment. Methods: A retrospective cohort study was conducted of a clinical–administrative database, reflecting all cases of non-reperfused ischaemic stroke admitted to Spanish hospitals during the period 2008–2012. A predictive model based on logistic regression was developed on a training cohort and later validated by the “hold-out” method. Complementary machine learning techniques were also explored. Results: The resulting model had the following nine variables, all readily obtainable during initial care. Age (OR 1.069), female sex (OR 1.202), readmission (OR 2.008), hypertension (OR 0.726), diabetes (OR 1.105), atrial fibrillation (OR 1.537), dyslipidaemia (0.638), heart failure (OR 1.518) and neurological symptoms suggestive of posterior fossa involvement (OR 2.639). The predictability was moderate (AUC 0.742, 95% CI: 0.737–0.747), with good visual calibration; Pearson’s chi-square test revealed non-significant calibration. An easily consulted risk score was prepared. Conclusions: It is possible to create a predictive model of mortality for patients with ischaemic stroke from which important advances can be made towards optimising the quality and efficiency of care. The model results are available within a few minutes of admission and would provide a valuable complementary resource for the neurologist.

Identification of vital modules and genes associated with heart failure based on weighted gene coexpression network analysis

Aims

Heart failure (HF) is a chronic heart disease with a high incidence and mortality. Due to the regulatory complexity of gene coexpression networks, the underlying hub genes regulation in HF remain incompletely appreciated. We aimed to explore potential key modules and genes for HF using weighted gene coexpression network analysis (WGCNA).

Methods and results

The expression profiles by high throughput sequencing of heart tissues samples from HF and non‐HF samples were obtained from the Gene Expression Omnibus database. Differentially expressed genes (DEGs) between HF and non‐HF samples were firstly identified. Then, a coexpression network was constructed to identify key modules and potential hub genes. The biological functions of potential hub genes were analysed by Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses. Finally, a protein–protein interaction (PPI) network was constructed using the STRING online tool. A total of 135 DEGs (133 up‐regulated and 2 down‐regulated DEGs) between HF and non‐HF samples were identified in the GSE135055 and GSE123976 datasets. Moreover, a total of 38 modules were screened based on WGCNA in the GSE135055 dataset, and six potential hub genes (UCK2, ASB1, CCNI, CUX1, IRX6, and STX16) were screened from the key module by setting the gene significance over 0.2 and the module membership over 0.8. Furthermore, 78 potential hub genes were obtained by taking the intersection of the 135 DEGs and all genes in the key module, and enrichment analysis revealed that they were mainly involved in the MAPK and PI3K‐AKT signalling pathways. Finally, in a PPI network constructed with the 78 potential hub genes, CUX1 and ASB1 were identified as hub genes in HF because they were also identified as potential hub genes in the WGCNA.

Conclusions

To the best of our knowledge, our study is the first to employ WGCNA to identify the key module and hub genes for HF. Our study identified a module and two genes that might play important roles in HF, which may provide potential biomarkers for the diagnosis of HF and improve our knowledge of the molecular mechanisms underlying HF.

The Comprehensive Machine Learning Analytics for Heart Failure

Background: Early detection of heart failure is the basis for better medical treatment and prognosis. Over the last decades, both prevalence and incidence rates of heart failure have increased worldwide, resulting in a significant global public health issue. However, an early diagnosis is not an easy task because symptoms of heart failure are usually non-specific. Therefore, this study aims to develop a risk prediction model for incident heart failure through a machine learning-based predictive model. Although African Americans have a higher risk of incident heart failure among all populations, few studies have developed a heart failure risk prediction model for African Americans. Methods: This research implemented the Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression, support vector machine, random forest, and Extreme Gradient Boosting (XGBoost) to establish the Jackson Heart Study's predictive model. In the analysis of real data, missing data are problematic when building a predictive model. Here, we evaluate predictors' inclusion with various missing rates and different missing imputation strategies to discover the optimal analytics. Results: According to hundreds of models that we examined, the best predictive model was the XGBoost that included variables with a missing rate of less than 30 percent, and we imputed missing values by non-parametric random forest imputation. The optimal XGBoost machine demonstrated an Area Under Curve (AUC) of 0.8409 to predict heart failure for the Jackson Heart Study. Conclusion: This research identifies variations of diabetes medication as the most crucial risk factor for heart failure compared to the complete cases approach that failed to discover this phenomenon.

A novel machine learning strategy for model selections - Stepwise Support Vector Machine (StepSVM)

An essential aspect of medical research is the prediction for a health outcome and the scientific identification of important factors. As a result, numerous methods were developed for model selections in recent years. In the era of big data, machine learning has been broadly adopted for data analysis. In particular, the Support Vector Machine (SVM) has an excellent performance in classifications and predictions with the high-dimensional data. In this research, a novel model selection strategy is carried out, named as the Stepwise Support Vector Machine (StepSVM). The new strategy is based on the SVM to conduct a modified stepwise selection, where the tuning parameter could be determined by 10-fold cross-validation that minimizes the mean squared error. Two popular methods, the conventional stepwise logistic regression model and the SVM Recursive Feature Elimination (SVM-RFE), were compared to the StepSVM. The Stability and accuracy of the three strategies were evaluated by simulation studies with a complex hierarchical structure. Up to five variables were selected to predict the dichotomous cancer remission of a lung cancer patient. Regarding the stepwise logistic regression, the mean of the C-statistic was 69.19%. The overall accuracy of the SVM-RFE was estimated at 70.62%. In contrast, the StepSVM provided the highest prediction accuracy of 80.57%. Although the StepSVM is more time consuming, it is more consistent and outperforms the other two methods.