Clinical and serum-based markers are associated with death within 1 year of de novo implant in primary prevention ICD recipients

Effect of implantable cardioverter defibrillator on primary prevention of sudden cardiac death in high-risk patients

OBJECTIVE

To investigate the effect of implantable cardioverter defibrillator (ICD) on primary prevention of sudden cardiac death (SCD) in patients with high risk.

METHODS

This retrospective analysis included 70 patients who received primary prevention of SCD by ICD implantation in Huzhou Central Hospital from March 2016 to May 2019. Based on survival, 15 patients who died during follow-up were placed into the death group and the 55 patients who survived were set as the survival group. The two groups were compared in terms of sex, age, non-sustained ventricular tachycardia (VT), diastolic pressure, systolic pressure, left ventricular ejection fraction (LVEF), urea nitrogen, serum creatinine, history of diabetes, history of atrial fibrillation, history of myocardial ischemia, history of dilated cardiomyopathy, history of hypertrophic cardiomyopathy, type I Brugada wave and cardiac function classification. Further, we analyzed the proportion of discharge, the survival of patients (Kaplan Meier method), and the risk factors of patient death (Logistic regression).

RESULTS

The analysis of baseline data showed that patients in the death group had older age and higher level of serum creatinine than the survival group (P<0.05), and the number of patients with non-sustained VT≥5 times/24 h in the survival group was higher than that in the death group (P<0.05). There was no obvious difference in other baseline indexes between the two groups (P>0.05). In addition, there was no difference in the proportion of patients receiving appropriate/inappropriate discharge (P>0.05) between the two groups. Follow-up data showed that 15 cases (21.43%) of spontaneous VT/ventricular fibrillation events were correctly diagnosed by pacemakers and properly treated by ICD (discharge or antitachycardia pacing (ATP)), while 55 cases (78.57%) received inappropriate ICD treatment. There were 15 patients (21.43%) who died during follow up, including 6 cases of cardiac insufficiency, 1 case of SCD, 2 cases of acute myocardial infarction, 1 case of respiratory failure, and 5 cases of unknown etiology; the survival time was (20.27±7.06) months. Logistic regression analysis showed that age and serum creatinine were the risk factors of patient death.

CONCLUSION

Primary prevention with ICD implantation benefits SCD patients. Non persistent VT≥5 times/24 h is a predictive value for ICD implantation in patients receiving primary prevention of SCD. Age and serum creatinine are risk factors for death.

Influence of diabetes on mortality and ICD therapies in ICD recipients: a systematic review and meta-analysis of 162,780 patients

Background

The influence of diabetes on the mortality and risk of implantable cardioverter defibrillator (ICD) therapies is still controversial, and a comprehensive assessment is lacking. We performed this systematic review and meta-analysis to address this controversy.

Methods

We systematically searched the PubMed, Embase, Web of Science and Cochrane Library databases to collect relevant literature. Fixed and random effects models were used to estimate the hazard ratio (HR) with 95% CIs.

Results

Thirty-six articles reporting on 162,780 ICD recipients were included in this analysis. Compared with nondiabetic ICD recipients, diabetic ICD recipients had higher all-cause mortality (HR = 1.45, 95% CI 1.36–1.55). The subgroup analysis showed that secondary prevention patients with diabetes may suffer a higher risk of all-cause mortality (HR = 1.89, 95% CI 1.56–2.28) (for subgroup analysis, P = 0.03). Cardiac mortality was also higher in ICD recipients with diabetes (HR = 1.68, 95% CI 1.35–2.08). However, diabetes had no significant effect on the risks of ICD therapies, including appropriate or inappropriate therapy, appropriate or inappropriate shock and appropriate anti-tachycardia pacing (ATP). Diabetes was associated with a decreased risk of inappropriate ATP (HR = 0.56, 95% CI 0.39–0.79).

Conclusion

Diabetes is associated with an increased risk of mortality in ICD recipients, especially in the secondary prevention patients, but does not significantly influence the risks of ICD therapies, indicating that the increased mortality of ICD recipients with diabetes may not be caused by arrhythmias. The survival benefits of ICD treatment in diabetes patients are limited.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12933-022-01580-y.

Machine Learned Cellular Phenotypes in Cardiomyopathy Predict Sudden Death

RATIONALE

Susceptibility to VT/VF (ventricular tachycardia/fibrillation) is difficult to predict in patients with ischemic cardiomyopathy either by clinical tools or by attempting to translate cellular mechanisms to the bedside.

OBJECTIVE

To develop computational phenotypes of patients with ischemic cardiomyopathy, by training then interpreting machine learning of ventricular monophasic action potentials (MAPs) to reveal phenotypes that predict long-term outcomes.

METHODS AND RESULTS

We recorded 5706 ventricular MAPs in 42 patients with coronary artery disease and left ventricular ejection fraction ≤40% during steady-state pacing. Patients were randomly allocated to independent training and testing cohorts in a 70:30 ratio, repeated K=10-fold. Support vector machines and convolutional neural networks were trained to 2 end points: (1) sustained VT/VF or (2) mortality at 3 years. Support vector machines provided superior classification. For patient-level predictions, we computed personalized MAP scores as the proportion of MAP beats predicting each end point. Patient-level predictions in independent test cohorts yielded c-statistics of 0.90 for sustained VT/VF (95% CI, 0.76-1.00) and 0.91 for mortality (95% CI, 0.83-1.00) and were the most significant multivariate predictors. Interpreting trained support vector machine revealed MAP morphologies that, using in silico modeling, revealed higher L-type calcium current or sodium-calcium exchanger as predominant phenotypes for VT/VF.

CONCLUSIONS

Machine learning of action potential recordings in patients revealed novel phenotypes for long-term outcomes in ischemic cardiomyopathy. Such computational phenotypes provide an approach which may reveal cellular mechanisms for clinical outcomes and could be applied to other conditions.

Cause-specific death and risk factors of one-year mortality after implantable cardioverter-defibrillator implantation: a nationwide study

AIMS

Current treatment guidelines recommend implantable cardioverter-defibrillators (ICDs) in eligible patients with an estimated survival beyond one year. There is still an unmet need to identify patients who are unlikely to benefit from an ICD.We determined cause-specific one-year mortality after ICD implantation and identified associated risk factors.

METHODS AND RESULTS

Using Danish nationwide registries (2000-2017), we identified 14,516 patients undergoing first-time ICD implantation for primary or secondary prevention. Risk factors associated with one-year mortality were evaluated using multivariable logistic regression. The median age was 66 years, 81.3% were male, and 50.3% received an ICD for secondary prevention. The one-year mortality rate was 4.8% (694/14,516). ICD recipients who died within one year were older and more comorbid compared to those who survived (72 vs. 66 years, p < 0.001). Risk factors associated with increased one-year mortality included dialysis (OR:3.26, CI:2.37-4.49), chronic renal disease (OR:2.14, CI:1.66-2.76), cancer (OR:1.51, CI:1.15-1.99), age 70-79 years (OR:1.65, CI:1.36-2.01), and age ≥80 years (OR:2.84, CI:2.15-3.77). The one-year mortality rates for the specific risk factors were: dialysis (13.8%), chronic renal disease (13.1%), cancer (8.5%), age 70-79 years (6.9%), and age ≥80 years (11.0%). Overall, the most common causes of mortality were related to cardiovascular diseases (62.5%), cancer (10.1%), and endocrine disorders (5.0%). However, the most common cause of death among patients with cancer was cancer-related (45.7%).

CONCLUSION

Among ICD recipients, mortality rates were low and could be indicative of relevant patient selection. Important risk factors of increased one-year mortality included dialysis, chronic renal disease, cancer, and advanced age.

Improving Clinical Translation of Machine Learning Approaches Through Clinician-Tailored Visual Displays of Black Box Algorithms: Development and Validation

BACKGROUND

Despite the promise of machine learning (ML) to inform individualized medical care, the clinical utility of ML in medicine has been limited by the minimal interpretability and black box nature of these algorithms.

OBJECTIVE

The study aimed to demonstrate a general and simple framework for generating clinically relevant and interpretable visualizations of black box predictions to aid in the clinical translation of ML.

METHODS

To obtain improved transparency of ML, simplified models and visual displays can be generated using common methods from clinical practice such as decision trees and effect plots. We illustrated the approach based on postprocessing of ML predictions, in this case random forest predictions, and applied the method to data from the Left Ventricular (LV) Structural Predictors of Sudden Cardiac Death (SCD) Registry for individualized risk prediction of SCD, a leading cause of death.

RESULTS

With the LV Structural Predictors of SCD Registry data, SCD risk predictions are obtained from a random forest algorithm that identifies the most important predictors, nonlinearities, and interactions among a large number of variables while naturally accounting for missing data. The black box predictions are postprocessed using classification and regression trees into a clinically relevant and interpretable visualization. The method also quantifies the relative importance of an individual or a combination of predictors. Several risk factors (heart failure hospitalization, cardiac magnetic resonance imaging indices, and serum concentration of systemic inflammation) can be clearly visualized as branch points of a decision tree to discriminate between low-, intermediate-, and high-risk patients.

CONCLUSIONS

Through a clinically important example, we illustrate a general and simple approach to increase the clinical translation of ML through clinician-tailored visual displays of results from black box algorithms. We illustrate this general model-agnostic framework by applying it to SCD risk prediction. Although we illustrate the methods using SCD prediction with random forest, the methods presented are applicable more broadly to improving the clinical translation of ML, regardless of the specific ML algorithm or clinical application. As any trained predictive model can be summarized in this manner to a prespecified level of precision, we encourage the use of simplified visual displays as an adjunct to the complex predictive model. Overall, this framework can allow clinicians to peek inside the black box and develop a deeper understanding of the most important features from a model to gain trust in the predictions and confidence in applying them to clinical care.

Clinical risk prediction with random forests for survival, longitudinal, and multivariate (RF-SLAM) data analysis

Background

Clinical research and medical practice can be advanced through the prediction of an individual’s health state, trajectory, and responses to treatments. However, the majority of current clinical risk prediction models are based on regression approaches or machine learning algorithms that are static, rather than dynamic. To benefit from the increasing emergence of large, heterogeneous data sets, such as electronic health records (EHRs), novel tools to support improved clinical decision making through methods for individual-level risk prediction that can handle multiple variables, their interactions, and time-varying values are necessary.

Methods

We introduce a novel dynamic approach to clinical risk prediction for survival, longitudinal, and multivariate (SLAM) outcomes, called random forest for SLAM data analysis (RF-SLAM). RF-SLAM is a continuous-time, random forest method for survival analysis that combines the strengths of existing statistical and machine learning methods to produce individualized Bayes estimates of piecewise-constant hazard rates. We also present a method-agnostic approach for time-varying evaluation of model performance.

Results

We derive and illustrate the method by predicting sudden cardiac arrest (SCA) in the Left Ventricular Structural (LV) Predictors of Sudden Cardiac Death (SCD) Registry. We demonstrate superior performance relative to standard random forest methods for survival data. We illustrate the importance of the number of preceding heart failure hospitalizations as a time-dependent predictor in SCA risk assessment.

Conclusions

RF-SLAM is a novel statistical and machine learning method that improves risk prediction by incorporating time-varying information and accommodating a large number of predictors, their interactions, and missing values. RF-SLAM is designed to easily extend to simultaneous predictions of multiple, possibly competing, events and/or repeated measurements of discrete or continuous variables over time.Trial registration: LV Structural Predictors of SCD Registry (clinicaltrials.gov, NCT01076660), retrospectively registered 25 February 2010

Implantable Cardioverter-Defibrillator Placement for Primary Prevention in 2,346 Patients: Predictors of One-Year Survival

Guidelines suggest that patients who receive implantable cardioverter-defibrillators (ICDs) for primary prevention should be expected to live more than one year after placement. However, tools for validating this prognosis are not sufficiently predictive. We sought to identify definitive predictors of one-year survival after ICD placement. By reviewing medical records and the Social Security Death Index, we analyzed baseline characteristics and survival outcomes of 3,164 patients who underwent ICD placement at our institution from January 2006 through March 2014. Survival outcome could be confirmed for 2,346 patients (74%). Of these, 184 (7.8%) died within one year of ICD placement. We noted significant differences in numerous variables between those who lived and died. However, multivariable analysis revealed only 5 independent predictors of earlier death: worse New York Heart Association functional class (hazard ratio [HR]=1.87 per class [95% CI, 1.22-2.87]; P <0.01); lower serum sodium level (HR=0.93 per 1 mEq/L increase [95% CI, 0.88-0.99]; P=0.04); atrial fibrillation (HR=1.81 [95% CI, 1.03-3.21]; P=0.04); chronic lung disease (HR=2.05 [95% CI, 1.20-3.51]; P <0.01), and amiodarone use (HR=10.1 [95% CI, 4.51-22.5]; P <0.01). Using receiver operating characteristic curves, we developed a model with an area under the curve of 0.718 that predicted death at one year after ICD implantation. Despite significant univariate differences between the ICD recipients who did and did not live beyond one year, we found only moderate predictors of survival. Better tools are needed to predict outcomes when considering ICD placement for primary prevention.

Improving sudden cardiac death risk stratification by evaluating electrocardiographic measures of global electrical heterogeneity and clinical outcomes among patients with implantable cardioverter-defibrillators: rationale and design for a retrospective, multicenter, cohort study

PURPOSE

Implantable cardioverter-defibrillators (ICDs) improve survival of systolic heart failure (HF) patients who are at risk of sudden cardiac death (SCD). We recently showed that electrocardiographic (ECG) global electrical heterogeneity (GEH) is independently associated with SCD in the community-dwelling cohort and developed GEH SCD risk score. The Global Electrical Heterogeneity and Clinical Outcomes (GEHCO) study is a retrospective multicenter cohort designed with two goals: (1) validate an independent association of ECG GEH with sustained ventricular tachyarrhythmias and appropriate ICD therapies and (2) validate GEH ECG risk score for prediction of sustained ventricular tachyarrhythmias and appropriate ICD therapies in systolic HF patients with primary prevention ICD.

METHODS

All records of primary prevention ICD recipients with available data for analysis are eligible for inclusion. Records of ICD implantation in patients with inherited channelopathies and cardiomyopathies are excluded. Raw digital 12-lead pre-implant ECGs will be used to measure GEH (spatial QRST angle, spatial ventricular gradient magnitude, azimuth, and elevation, and sum absolute QRST integral). The primary endpoint is defined as a sustained ventricular tachyarrhythmia event with appropriate ICD therapy. All-cause death without preceding sustained ventricular tachyarrhythmia with appropriate ICD therapy will serve as a primary competing outcome. The study will draw data from the academic medical centers.

RESULTS

We describe the study protocol of the first multicenter retrospective cohort of primary prevention ICD patients with recorded at baseline digital 12-lead ECG.

CONCLUSION

Findings from this study will inform future trials to identify patients who are most likely to benefit from primary prevention ICD.

TRIAL REGISTRATION

URL: http://www.clinicaltrials.gov . Unique identifier: NCT03210883.

New York Heart Association class and the survival benefit from primary prevention implantable cardioverter defibrillators: A pooled analysis of 4 randomized controlled trials

BACKGROUND

Primary prevention implantable cardioverter defibrillator (ICD) reduce all-cause mortality by reducing sudden cardiac death. There are conflicting data regarding whether patients with more advanced heart failure derive ICD benefit owing to the competing risk of nonsudden death.

METHODS

We performed a patient-level meta-analysis of New York Heart Association (NYHA) class II/III heart failure patients (left ventricular ejection fraction ≤35%) from 4 primary prevention ICD trials (MADIT-I, MADIT-II, DEFINITE, SCD-HeFT). Bayesian-Weibull survival regression models were used to assess the impact of NYHA class on the relationship between ICD use and mortality.

RESULTS

Of the 2,763 patients who met study criteria, 68% (n=1,867) were NYHA II and 52% (n=1,435) were randomized to an ICD. In a multivariable model including all study patients, the ICD reduced mortality (hazard ratio [HR] 0.65, 95% posterior credibility interval [PCI]) 0.40-0.99). The interaction between NYHA class and the ICD on mortality was significant (posterior probability of no interaction=.036). In models including an interaction term for the NYHA class and ICD, the ICD reduced mortality among NYHA class II patients (HR 0.55, PCI 0.35-0.85), and the point estimate suggested reduced mortality in NYHA class III patients (HR 0.76, PCI 0.48-1.24), although this was not statistically significant.

CONCLUSIONS

Primary prevention ICDs reduce mortality in NYHA class II patients and trend toward reducing mortality in the heterogeneous group of NYHA class III patients. Improved risk stratification tools are required to guide patient selection and shared decision making among NYHA class III primary prevention ICD candidates.

Improving the appropriateness of sudden arrhythmic death primary prevention by implantable cardioverter-defibrillator therapy in patients with low left ventricular ejection fraction. Point of view

It is generally accepted that the current guidelines for the primary prevention of sudden arrhythmic death, which are based on ejection fraction, do not allow the optimal selection of patients with low left ventricular ejection fraction of ischemic and nonischemic etiology for implantation of a cardioverter-defibrillator. Ejection fraction alone is limited in both sensitivity and specificity. An analysis of the risk of sudden arrhythmic death with a combination of multiple tests (ejection fraction associated with one or more arrhythmic risk markers) could partially compensate for these limitations. We propose a polyparametric approach for defining the risk of sudden arrhythmic death using ejection fraction in combination with other clinical and arrhythmic risk markers (i.e. late gadolinium enhancement cardiac magnetic resonance, T-wave alternans, programmed ventricular stimulation, autonomic tone, and genetic testing) that have been validated in nonrandomized trials. In this article, we examine these approaches to identify three subsets of patients who cannot be comprehensively assessed by the current guidelines: patients with ejection fraction of 35% or less and a relatively low risk of sudden arrhythmic death despite the ejection fraction value; patients with ejection fraction of 35% or less and high competitive risk of death due to evolution of heart failure or noncardiac causes; and patients with ejection fraction between 35 and 45% with relatively high risk of sudden arrhythmic death despite the ejection fraction value.

Troponin levels after ICD implantation with and without defibrillation testing and their predictive value for outcomes: Insights from the SIMPLE trial

BACKGROUND

The Shockless IMPLant Evaluation trial randomized 2500 patients receiving a first implantable cardioverter-defibrillator (ICD)/cardiac resynchronization therapy-defibrillator device to have either defibrillation testing (DT) or no DT. It demonstrated that DT did not improve shock efficacy or reduce mortality.

OBJECTIVE

This prospective substudy evaluated the effect of DT on postoperative troponin levels and their predictive value for total and arrhythmic mortality.

METHODS

Troponin levels were measured between 6 and 24 hours after ICD implantation in 2200 of 2500 patients.

RESULTS

A postoperative serum troponin level above the upper limit of normal (ULN) was more common in patients undergoing DT (n = 509 [46.4%]) than in those not subjected to DT (n = 456 [41.3%]; P = .02). After excluding patients with known preoperative troponin levels above the ULN, consistent findings were observed (42.1% vs 37.5%; P = .04). During a mean follow-up of 3.1 ± 1.0 years, the annual mortality rate was increased in patients with postoperative troponin levels above the ULN (adjusted hazard ratio [HR] 1.43; 95% confidence interval [CI] 1.15-1.76; P = .001) irrespective of DT or no DT. Likewise, patients with elevated troponin levels had a significantly higher risk of arrhythmic death (adjusted HR 1.80; 95% CI 1.23-2.63; P = .002). The rate of first appropriate ICD shock (adjusted HR 0.89; 95% CI 0.71-1.12; P = .32) or failed appropriate shock (adjusted HR 1.02; 95% CI 0.59-1.76; P = .95) was similar in patients with or without troponin elevation.

CONCLUSION

DT at the time of ICD implantation is associated with increased troponin levels, indicating subclinical myocardial injury caused by the procedure. Elevated troponin levels but not DT seem to predict clinical outcomes in ICD recipients.