Citations

Abstract Purpose While randomized controlled trials remain the gold standard for assessing treatment efficacy, studies using real-world data (RWD) offer valuable insights into treatment effectiveness across broader, more diverse patient populations. This commentary explores the importance of using fit-for-purpose data and emphasizes the need for rigorous evaluation of RWD quality to support valid and actionable evidence generation. Summary The utility of RWD-based studies depends heavily on the fitness-for-purpose of the data source, which requires careful assessment of 5 key quality dimensions: relevance, extensiveness, timeliness, coherence, and reliability. Practical examples from coronavirus disease 2019 (COVID-19) comparative effectiveness research are used to illustrate each data quality domain. Conclusions As the need for RWD increases, especially for post–COVID-19 pandemic decision-making, ensuring high data quality and appropriate study design is critical. Proper evaluation of RWD sources enhances the credibility of findings and supports their use in meta-analyses, clinical guidelines, and healthcare policy.

Abstract Motivation The American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines represent the gold standard for clinical variant interpretation. Despite the widespread adoption of ACMG/AMP guidelines, a comprehensive comparison of the software tools designed to implement them has been lacking. This represents a significant gap, as clinicians require evidence-based guidance on which tools to use in their practice. Results We benchmarked four ACMG/AMP-based tools (Franklin, InterVar, TAPES, Genebe) selected from 22 tools, and compared their performance with LIRICAL, a top-performing phenotype-driven tool, using 151 expert-curated datasets from Mendelian disorders. Selection criteria included free availability, VCF compatibility, operational reliability, and not being disease-specific. Our evaluation framework assessed top-N accuracy (N = 1, 5, 10, 20, 50), retention rates, precision, recall, F1 scores, and area under the curve (AUC). Statistical validation employed bootstrap confidence intervals (n = 1000) and Friedman tests. LIRICAL (68.21%) and Franklin (61.59%) demonstrated superior top-10 variant prioritization accuracy in Mendelian disorders, significantly outperforming other tools (P = .0000). Results demonstrate that tools with advanced phenotypic integration significantly outperform those relying primarily on genomic features. Availability and implementation All data and source code required to reproduce the findings of this study are openly available in the Code Ocean repository at https://doi.org/10.24433/CO.6562438.v1.

Abstract Background Inducible ST-segment depression during stress echocardiography (SE) is associated with a higher risk in patients with inducible regional wall motion abnormality (RWMA). Objective To assess ST segment elevation's prevalence and functional correlates during SE. Methods ECG response was analyzed in 1712 patients (age 63±12 years; 1011 men, 59%) undergoing SE for chronic coronary syndromes (CCS) prospectively recruited at 39 laboratories. Among the patients analyzed, 1286 (75%) performed SE with exercise, 306 (18%) with dobutamine and 120 (7%) with a vasodilator. ECG response was considered abnormal in case of ST-segment shift >1.0 mm from baseline at 80 ms from the J point developed in ≥2 contiguous leads during SE. The stress to rest (∆) change in the wall motion score index (WMSI) was assessed in all patients. Coronary angiography was available in 765 patients. Major adverse cardiac events (MACE) were the outcome measure in 746 patients with follow-up information. Results No significant ST segment shift was observed in 1000 patients (58%, Group 1), ST-segment depression occurred in 609 patients (35.5%, Group 2), and ST-segment elevation in 103 patients (6%, Group 3). ST-segment elevation occurred in 66/1286 patients (5.2%) with exercise, 30/306 with dobutamine (10%), and 7/120 with vasodilators (6%, p = ns vs. exercise and p<0.01 vs. dobutamine). The prevalence of stress-induced RMWA was higher in Group 2 (38.1%) compared to Group 3 (33%) and Group 1 (11.9%. P< 0.001 vs. Groups 2 and 3); however, the ΔWMSI was more pronounced in Group 3 compared to Group 2 and Group 1 (p = ns vs. Group 2 and p<0.001 vs. Group 1). (Figure, left panel). Non-significant CAD (defined as coronary stenosis <50% on coronary angiography) was more prevalent in Group 1 (36%) and 3 (40%) compared to Group 2 (17.5%, p<0.01 vs. Group 1 and 3). . During a median follow-up of 387 days (IQ range 314-603 days), 150 events occurred in 112 patients: 14 all-cause deaths, 39 acute myocardial infarctions, 19 atrial fibrillations, 34 late coronary revascularizations, and 44 acute heart failure hospitalizations. The exposure-adjusted event rate was the lowest in Group 1 compared to Groups 2 and 3 (Figure, right panel). Conclusion In all-comers with CCS, the electrocardiographic pattern of ST-segment elevation is not rare, more frequent with dobutamine, associated with more stress-to-rest change in RWMA and more events despite less extensive underlying anatomic CAD. The paradox of more ischemia with less coronary stenosis is consistent with a coronary vasospastic origin as an important yet frequently overlooked cause of stress-induced ST segment elevation.Figure

Abstract Background Cardiovascular disease (CVD) is the leading cause of death in patients with type 2 diabetes mellitus (T2DM). Glucagon-like peptide-1 receptor agonists (GLP-1RAs) and sodium-glucose cotransporter 2 inhibitors (SGLT2i) have demonstrated cardiovascular (CV) benefits and are recommended for T2DM patients with atherosclerotic CVD (ASCVD). However, real-world data on their adoption in high-risk populations, particularly myocardial infarction (MI) and stroke survivors, remain limited. Purpose To assess real-world utilization of GLP-1RAs and SGLT2i among T2DM patients with ASCVD focusing on trends, disparities by age, sex, and region from 2015 to 2023. Methods A nationwide cohort study using data from the National Registry of Reimbursed Health Services (2015–2023), covering all reimbursed health services in the Czech Republic. Patients with nonfatal MI (ICD-10: I21, I22) or stroke (I63, I64) and confirmed T2DM were included. Prescriptions were identified at the event or within 12 months post-event. Utilization trends were analyzed, and logistic regression was used to determine predictors of medication use. Results Among 116,218 MI patients (median age:68 years, 33.5% women), 53,128 (46%) had T2DM (median age:71 years, 37% women). Within 12 months post-MI, 2.1% received GLP-1RAs and 10.6% received SGLT2i, with increasing trends from 0.3% (2015) to 4.8% (2023) (GLP-1RAs) and 0.4% (2015) to 34.8% (2023) (SGLT2i). Women were less likely to receive SGLT2i (aOR: 0.70 [95%CI:0.65–0.74]), while no significant difference was observed for GLP-1RAs (aOR:0.98 [95%CI:0.86–1.12]). Older patients (>70 years) had lower odds of receiving GLP-1RAs (aOR:0.31 [95%CI:0.27–0.36]) and SGLT2i (aOR:0.53 [95%CI:0.50–0.56]). Prescription rates were highest in Prague (3.1% GLP-1RAs; 13.5% SGLT2i) and lowest in the Moravian-Silesian Region (1.4% GLP-1RAs; 8.8% SGLT2i). Among 164,729 stroke patients (median age:73 years, 49% women), 78,562 (48%) had T2DM (median age:75 years, 50% women). Within 12 months post-stroke, 1.2% received GLP-1RAs and 3.7% received SGLT2i, with increasing trends from 0.2% (2015) to 2.8% (2023) (GLP-1RAs) and 0.1% (2015) to 10.4% (2023) (SGLT2i). Women were less likely to receive GLP-1RAs (aOR:0.82 [95%CI:0.72–0.95]) and SGLT2i (aOR:0.57 [95%CI:0.52–0.62]). Older patients (>70 years) had lower odds of receiving GLP-1RAs (aOR:0.31 [95%CI:0.27–0.36]) and SGLT2i (aOR:0.40 [95%CI:0.37–0.44]). Regional disparities were notable, with prescription rates varying from 0.8–2.1% (GLP-1RAs) and 2.4–5.2% (SGLT2i). Conclusions Prescribing rates for GLP-1RAs and SGLT2i have increased over the past eight years but remain low, even among MI and stroke survivors. Women, older patients, and those in lower-income regions are significantly less likely to receive these therapies. Targeted interventions are needed to improve secondary prevention in high-risk diabetic populations and to address these disparities.Myocardial infarction patients Stroke patients

Abstract Background Ventricular systolic dysfunction (LVSD) is the main predictor of mortality in Chagas disease (ChD). Although LVSD can be treated with affordable medications that enhance both symptoms and survival, its diagnosis necessitates advanced imaging tests, which are often unavailable in resource-limited settings. Recently, artificial intelligence (AI) models applied to an electrocardiogram (ECG) have shown promise in detecting LVSD and assessing disease progression, acting as digital biomarkers and offering the potential for screening tools. Purpose This study aims to evaluate the performance of an AI-ECG model in detecting LVSD and predicting mortality in patients with ChD. Specifically, we will fine-tune the AI-ECG model to improve LVSD identification and update the previously published mortality score (JAHA) by replacing NT-proBNP with AI-ECG to assess mortality risk. Methods This study was conducted with patients from the SaMi-Trop project, a prospective cohort of individuals seropositive for ChD from 21 municipalities in Minas Gerais, an endemic region of Brazil. The baseline assessment occurred in 2013–2014, with follow-ups in 2015–2016 (FU1) and 2022–2023 (FU2). The AI-ECG model was fine-tuned using a dataset with an 8% low ejection fraction prevalence. The model was trained with a binary cross-entropy loss function, a batch size of 128, and an Adam optimizer with a learning rate of 0.001. The area under the curve (AUC) was calculated for validation. The fine-tuned model, parsed into 2-second ECG windows, was applied to two waves of the SaMi-Trop cohort: Baseline (4,701 ECGs from 1,906 patients) and FU1 (2,324 ECGs from 1,059 patients). The association between AI-detected LVSD (AI-LVSD) and mortality was assessed using Cox proportional hazard models. Results The fine-tuned AI-LVSD model achieved an AUC of 0.88 (95% CI: 0.78–0.97) in detecting LVSD. When updating the JAHA mortality score, AI-LVSD replaced NT-proBNP with a stronger association with mortality risk, yielding hazard ratios (HR) of 4.5 (95% CI: 3,0–6.8) for two-year mortality and 4.3 (95% CI: 3.4–5.5) for overall mortality. Additionally, when incorporated into a simplified hazard model alongside age and NYHA class, AI-LVSD emerged as a strong biomarker for mortality risk, with hazard ratios of 5.4 (95% CI: 3.7–7.8) for two-year mortality and 4.9 (95% CI: 4.0–6.1) for overall mortality (Figure 1). Conclusion The AI-ECG model accurately identified LVSD for screening purposes and effectively replaced NT-proBNP in a simplified mortality risk model for Chagas disease. This tool can potentially enhance early detection and risk stratification, particularly in remote areas with limited access to advanced diagnostic resources.Figure 01.Cox proportional hazard model

Abstract Background Obese patients often exhibit a clustering of cardiovascular risk factors and are frequently referred for noninvasive cardiac imaging. Coronary microvascular dysfunction (CMD) represents an early indicator of cardiovascular disease that can be noninvasively diagnosed using coronary flow velocity reserve (CFVR) measurements obtained during stress echocardiography (SE) in patients without evidence of obstructive coronary artery disease (CAD). Purpose This study aimed to evaluate the prognostic significance of CFVR in overweight and obese patients within a prospective multicenter cohort. Methods 2273 patients (mean age 65 ± 11 years; 51% women) classified as overweight or obese (for definition see Table) were included in 27 sites from 11 countries. All patients had normal left ventricular ejection fraction (EF ≥ 50%), no prior history of CAD, and underwent exercise or pharmacological SE, which was negative for inducible ischemia. CFVR of the left anterior descending coronary artery was assessed, with CMD defined as CFVR >2.0. Patients were divided into two groups based on the presence (Group 1) or absence (Group 2) of CMD. The primary endpoint was overall mortality. Results CMD was identified in 492 patients (22%), with a mean CFVR of 2.3 ± 0.5 across the study population. Group 1 patients (CMD) were older, with a greater prevalence of diabetes and metabolic syndrome compared to group 2 (no CMD) (Table). During a median follow-up period of 2.1 years (IQR: 1.2-3.7 years), 103 patients (5%) died. Patients with CMD had significantly worse late survival compared to those without CMD (3-year survival: 96 ± 1% vs. 96 ± 1%, p=ns; 6-year survival: 79 ± 4% vs. 93 ± 1%, p < 0.0001): see Figure. Multivariate analysis adjusting for age, sex, EF, body mass index, smoking habits, diabetes, dyslipidemia, arterial hypertension, and type of stress confirmed CMD as an independent predictor of mortality (hazard ratio: 1.8 [95% CI: 1.1–2.8], p = 0.01). Conclusions CMD, assessed via CFVR during SE, was a robust and independent prognostic marker of long-term mortality in obese patients with preserved EF and no evidence of inducible ischemia. These findings underscore the pivotal role of noninvasive assessment of CMD in refining risk stratification and guiding clinical decision-making in this high-risk population.

Abstract Background Computed tomography coronary angiography (CCTA) has emerged as a powerful non-invasive imaging modality for the evaluation of coronary artery disease (CAD), with a global trend towards a "CT-first" strategy in stable patients with suspected CAD. While overall beneficial, this approach requires a shift in resources, straining specific parts of healthcare systems. Pre-test probability (PTP) models stratify risks of obstructive CAD and are intended to result in better utilization of CCTA. However, these models show variability across subpopulations and have modest discriminatory power with low positive predictive values (PPV). Purpose This study aims to leverage explainable artificial intelligence (XAI) for optimization of PTP models. Methods Patients that underwent a CCTA for suspected CAD with a CAD-Reporting and Data System (CAD-RADS) classification and Coronary Artery Calcium Score (CACS) were included. Obstructive CAD was defined as stenosis >50% (CAD-RADS 3 or higher). Missing data was imputed. The Risk Factor Weighted Clinical Likelihood (RF-CL) model based on clinical risk factors (age, sex, type of symptoms, smoking, hypertension, dyslipidemia, diabetes and familial predisposition) was calculated. Two gradient descent-based decision tree models, "AI-ECG" and "AI-CAC" were developed. AI-ECG is built using clinical risk factors and electrocardiogram (ECG) classification. AI-CAC is built using the aforementioned features with addition of CACS. Sensitivity analysis was performed using a 15% threshold, aligning with the threshold for CCTA acquisition. Comparative analysis was performed using Python 3.10.2, with area under the curve (AUC) analysis using the DeLong test and feature importance calculation using shapely values. Preliminary results: We identified 6178 patients with a CCTA between 2022 and 2024 in the Cardiology Centers Netherlands (CCN) and included 2083 patients (mean age 57.4 ± 10.2; 48.1% male). Obstructive CAD was present in 413 (19.8%) patients. The RF-CL model was outperformed by the AI-ECG model (AUC: 0.69 vs 0.80, p<0.001) and the AI-CACS model (AUC: 0.69 vs 0.90, p<0.001). Although sensitivity increased in both the AI-ECG and AI-CACS models from 0.46 to 0.87 and 0.83 respectively, PPV increased only in the AI-CACS model (0.37 to 0.61). Most important features were CACS, age, sex, ECG classification and familial predisposition. Conclusion and Discussion Our study demonstrates that the incorporation of XAI with clinical risk factors, ECG classification and CACS improves the pre-test probability estimation of obstructive CAD. ECG classification appears to be a robust feature. Adopting these models can stratify patients at risk for obstructive CAD, enabling targeted use of CCTA, reducing unnecessary testing, and thus alleviate the burden on healthcare systems. However, further external validation is necessary to assess the model’s performance in prospective real-world settings.Sensitivity analysis feature importance

Abstract Background Loeys-Dietz Syndrome (LDS) is a connective tissue disorder caused by mutations in the TGFBR1 or TGFBR2 genes, leading to significant cardiovascular complications, particularly aortic aneurysms and dissections. Due to the progressive nature of these issues, patients often require multiple cardiac interventions throughout their lives, although this has not been comprehensively reviewed in the literature. Purpose This study aims to review trends in cardiac interventions in LDS patients to better understand the surgical course these patients go through in their lives. Methods A retrospective analysis was conducted on a cohort of individuals diagnosed with LDS at our institution. Medical records were reviewed for data on demographics and types of interventions (e.g., aortic surgeries, valve procedures) were collected. Descriptive statistics were used to analyze the frequency of interventions with comparisons made between intervention types. Results A total of 94 patients with LDS were included in the study. Among them, 69 patients (73.4%) had an ascending aortic aneurysm, and 13 patients (14%) had a type A aortic dissection. A total of 60 patients underwent aortic root replacement surgery due to aneurysm or dissection, with 10 patients (11%) having the procedure following the occurrence of a dissection. Mitral valve prolapse (MVP) was found in 34 patients (36%). Valvular interventions were performed as follows, with some patients undergoing repairs on multiple valves: 12 patients (13%) had mitral valve repair, 10 patients (11%) had tricuspid valve repair, and the aortic valve was the most intervened upon, with 37 patients (39.5%) undergoing either replacement or repair. In total, 45 patients (48%) underwent at least one valvular intervention. Additionally, 4 patients (4.5%) underwent ventricular septal defect (VSD) closure, 3 (3.5%) underwent atrial septal defect (ASD) closure, and 3 (3.5%) had both defects closed. 29 patients (31%) had a patent foramen ovale, with 17 of these patients (59%) undergoing PFO closure. Finally, 16 patients (17%) underwent coronary artery bypass grafting (CABG). Conclusion Patients with LDS often require multiple surgeries and interventions throughout their lives due to the progressive nature of the disease and its impact on the cardiovascular system. The high incidence of aortic aneurysms, dissections, and valvular abnormalities necessitates frequent surgical management, with many patients undergoing interventions such as aortic root replacement, valve repairs, and septal defect closures. Given the complex and multifaceted cardiovascular involvement in LDS, close monitoring and timely intervention are crucial for optimizing outcomes and improving quality of life for these patients.

Abstract Background Inability to reach age-predicted peak heart rate (APPHR) of >85%, or submaximal heart rates, on treadmill exercise stress testing (TMET) is associated with increased mortality risk [1]. Evidence is limited on mortality outcomes with heart rates greater than APPHR, or ‘supramaximal’ heart rates, especially when using heart rate metrics alone. Purpose Use of a comprehensive, compositive assessment of heart rate, termed ‘vigor’, to determine mortality outcomes has yet to be explored, especially for those with supramaximal heart rates. We sought to investigate this relationship. Methods We queried our stress database from 1993-2010 for patients >30 years old without cardiovascular disease and not on antichronotropic therapy. Clinical and mortality outcomes were retrieved from patient records. Patients were stratified into groups based on APPHR as achieved on TMET: <75%, 75-84%, 85-94%, 95-104%, ≥105% of APPHR. For our study, heart rate ‘vigor’ served as a composite assessment of cardiovascular fitness based solely on heart rate. This was derived from resting heart rate, heart rate recovery, and achieved percentage of APPHR. A higher Heart Rate Vigor Score (HRViS) for this study was, by definition, indicative of better cardiac responsiveness and recovery following exertion. Patients were stratified based on their achieved HRViS. Scores were determined as follows: All patients were initially assigned 3 points. 1 point was subtracted for each of the following parameters met: resting heart rate >90 beats per minute (BPM), heart rate recovery <13 BPM, or submaximal heart rates of APPHR <85%. Patients that remained at HRViS 3 and achieved supramaximal heart rates (≥105% APPHR) were assigned an additional point, amounting to HRViS 4. HRViS stratifications are detailed in Table 1. Mortality rates were calculated for all HRViS stratifications, with adjustment for FAC and known confounders. HRViS 3 was used as the referent. Results 18,961 total patients were included. 88 patients were classified with HRViS of 0; 1,763 as HRViS 1; 5,409 as HRViS 2; 9,902 as HRViS 3; and 1,799 as HRViS 4. Those with lower HRViS scores had more comorbid conditions (Table 1). 2,144 patients achieved submaximal APPHR (<85%) on TMET, and 2,917 achieved supramaximal APPHR (≥105%). HRViS 0, 1, and 2 had higher mortality risks compared to referent HRViS 3: 0 [3.76 (2.34-6.04), p<0.0001]; 1 [1.80 (1.53-2.12), p<0.0001]; 2 [1.29 (1.12-1.48), p=0.0004]. HRViS 4 had a non-statistically significant trends towards lower mortality risk compared to 3: 0.85 (0.66-1.08), p=0.1768. Conclusions HRViS 4 had a non-statistical trend towards lower mortality risk. Our findings do not support concern for adverse outcomes with excessive heart rates during exercise, and rather, may indicate a trend towards mortality benefit.HRViS Table 1