Daily Cardiology Research Analysis
Three cardiology studies stand out today: an AI-enabled ECG “sex discordance” score improved prediction of incident atrial fibrillation in women across multinational cohorts; a newly validated AL amyloidosis staging system (AL-ISS) incorporating longitudinal strain robustly identifies an ultra-poor risk stage (IIIC); and intravascular imaging-guided PCI in acute MI reduced 3-year MACE for complex (ACC/AHA B2/C) lesions.
Summary
Three cardiology studies stand out today: an AI-enabled ECG “sex discordance” score improved prediction of incident atrial fibrillation in women across multinational cohorts; a newly validated AL amyloidosis staging system (AL-ISS) incorporating longitudinal strain robustly identifies an ultra-poor risk stage (IIIC); and intravascular imaging-guided PCI in acute MI reduced 3-year MACE for complex (ACC/AHA B2/C) lesions.
Research Themes
- Risk stratification and predictive analytics in cardiology
- Imaging-guided intervention in acute coronary syndromes
- Cardio-oncology staging and prognostication
Selected Articles
1. Artificial intelligence-enabled electrocardiographic sex discordance and the risk of incident atrial fibrillation: a multi-national cohort study.
An AI-ECG model generated a continuous “sex discordance” score that was strongly associated with incident AF risk in women across three large test cohorts and improved CHARGE-AF discrimination. No significant association was seen in men; the score correlated with hormone imbalance, adiposity, and atrial remodeling in women.
Impact: This introduces a scalable AI-derived metric that refines AF risk prediction specifically in women, addressing sex-related risk heterogeneity and enabling targeted prevention.
Clinical Implications: Incorporating AI-ECG sex discordance into risk models may improve AF screening and risk factor modification strategies in women, informing monitoring intensity and lifestyle interventions.
Key Findings
- External validation AUCs for AI-ECG sex prediction were 0.91 (CODE-15%) and 0.90 (MIMIC-IV).
- Each SD increase in sex discordance score was associated with higher AF risk in females (HR 1.28 and 1.32 in two cohorts), but not in males.
- Adding the score to CHARGE-AF improved C-index in females by 0.026 and 0.020; replicated in UK Biobank (HR 1.23; C-index +0.024).
- In females, the score correlated with sex hormone imbalance, pericardial/visceral adiposity, atrial remodeling, and adverse lifestyle factors.
Methodological Strengths
- Multi-national external validation across large cohorts with consistent findings
- Improved discrimination beyond established CHARGE-AF risk model and mechanistic correlates
Limitations
- Observational design limits causal inference; model performance and thresholds require clinical implementation work
- Male cohorts showed no significant association, limiting generalizability across sexes
Future Directions: Prospective implementation studies to test screening strategies triggered by sex discordance in women, calibration across devices/populations, and integration with biomarkers and imaging.
BACKGROUND: Binary classification of sex fails to capture the sex-related continuum of atrial fibrillation (AF) risk. OBJECTIVE: To develop an artificial intelligence (AI)-enabled electrocardiography (ECG) model for sex prediction and explore its association with AF risk. METHODS: AI-ECG model for sex prediction was developed from Severance Hospital training set and externally validated using CODE-15% (AUC 0.91) and MIMIC-IV (AUC 0.90) datasets. A sex discordance score-defined as 1 minus the AI-ECG predicted probability (continuous) for self-reported sex-was estimated in AF-free individuals on three multi-national (Severance Hospital [n=205,769], Yongin Severance Hospital [n=112,942], and UK Biobank [n=40,525]) test sets. RESULTS: In Severance Hospital and Yongin Severance Hospital test set, sex discordance score increase was associated with higher AF risk in females, with CHARGE-AF-adjusted HR per SD of 1.28 (95% CI, 1.24-1.33) and 1.32 (95% CI, 1.27-1.36), respectively. No significant association was observed in males. Adding sex discordance score to CHARGE-AF model significantly improved discrimination for AF in females, with C-index increase of 0.026 (95% CI, 0.013-0.037) and 0.020 (95% CI, 0.010-0.032) in the respective datasets, but not in males. In UK Biobank test set, similar association between sex discordance score and incident AF risk was observed in females (CHARGE-AF-adjusted HR per SD, 1.23 [95% CI, 1.13-1.32]; C-index increase, 0.024 [95% CI, 0.005-0.040]). In females, sex discordance score correlated with sex hormone imbalance, pericardial and visceral adiposity, atrial remodeling, and adverse lifestyle factors. CONCLUSIONS: AI-ECG sex discordance score captures females with disproportionately elevated AF risk with implications for enhanced risk factor modification and surveillance.
2. A new validated staging system for AL amyloidosis with Stage IIIC defining ultra-poor risk: AL International Staging System (AL-ISS).
AL-ISS integrates longitudinal strain with NT-proBNP and troponin-T to stratify AL amyloidosis across five stages and robustly identifies an ultra-poor risk Stage IIIC with median survival of 7 months, validated across international cohorts in the modern treatment era.
Impact: The staging system modernizes risk stratification by adding strain imaging, delineates an ultra-poor risk group, and can immediately guide therapy intensity and trial design.
Clinical Implications: Incorporate LS into routine staging; identify Stage IIIC to prioritize aggressive therapy, referral to specialized centers, and consideration for novel/clinical trial regimens even in the daratumumab era.
Key Findings
- AL-ISS combines LS with NT-proBNP and hs-TnT to define stages I, II, IIIA, IIIB, and IIIC.
- In 2,493 patients, Stage IIIC had median survival of 7 months; IIIA 67 months; IIIB 26 months; I–II not reached.
- External validation showed good performance (12-month calibration slope 1.09; Harrell’s C 0.69).
- The ultra-poor risk IIIC stage remained prognostically adverse even among first-line daratumumab-treated patients (1-year OS 53% vs 68% for IIIB).
Methodological Strengths
- Large derivation and multinational external validation cohorts with contemporary treatments
- Use of calibration and discrimination metrics with clinically meaningful thresholds
Limitations
- Observational design; strain measurement variability across centers possible
- Moderate discrimination (Harrell’s C 0.69); treatment heterogeneity may confound outcomes
Future Directions: Prospective validation of AL-ISS-guided treatment algorithms, standardization of LS acquisition, and integration with genomic markers to refine risk and personalize therapy.
BACKGROUND: Outcomes in systemic AL amyloidosis have improved with modern therapy limiting utility of existing risk stratification models. We validate a new staging system, incorporating longitudinal strain (LS) to the biomarker-based (NT-proBNP and Troponin-T) staging system in the contemporary treatment era (2015-2024). METHODS: AL International Staging System (AL-ISS) was derived from a cohort of patients with AL amyloidosis from the UK National Amyloidosis Centre (2015-2019). The model was validated in patient cohorts from Europe (Greece, Italy, Netherlands, Switzerland), USA (2015-2024) and UK (2020-2024). RESULTS: 2493 patients were included (derivation, n=573; validation n=1920). In a multivariable model for the derivation cohort, LS≥-9% and cardiac biomarkers at previously validated thresholds (NT-proBNP 332 ng/L and 8500 ng/L and hs-TnT>50 ng/L) were independent poor prognostic factors stratifying patients into stages I, II, IIIA, IIIB and IIIC. In the validation cohort, the patient stages were stage I: 317 (17%), II: 782 (41%), IIIA: 551 (29%), IIIB: 174 (9%) and IIIC: 96 (5%), respectively (first-line daratumumab treated: 826; 43%). With a median follow-up of 34 months, median overall survival (OS) was not reached (NR); estimated 1-year, 2-year and 3-year OS was 82%, 74% and 70% respectively. Median survival for stages I-II, IIIA, IIIB and IIIC were NR, 67, 26 and 7 months (1-year OS IIIC 53% v 68% for IIIB in the daratumumab-treated patients), respectively (p<0.001). External validation exhibited good predictive performance: 12-month calibration slope was 1.09, Harrell's C 0.69, Royston's D 1.19, R CONCLUSION: This defines and validates a new staging system from systemic AL amyloidosis with robust identification of an ultra-poor risk stage (IIIC) in contemporarily treated patients.
3. Intravascular imaging-guided percutaneous coronary intervention for acute myocardial infarction according to ACC/AHA lesion classification.
Across 23,051 AMI patients from two national registries, IVI-guided PCI reduced 3-year MACE for complex (B2/C) lesions but not for simpler (A/B1) lesions. Benefits were consistent in both NSTEMI and STEMI subsets, with greater lesion complexity amplifying IVI’s prognostic advantage.
Impact: Findings support selective use of IVI in AMI based on lesion complexity, informing resource allocation and procedural strategy to improve outcomes.
Clinical Implications: Routine IVI use should be prioritized for ACC/AHA B2/C lesions in AMI to lower MACE; angiography alone may suffice for A/B1 lesions, enabling tailored imaging strategies.
Key Findings
- Patient-level pooled analysis (n=23,051) from KAMIR-V and KAMIR-NIH registries.
- IVI-guided PCI reduced 3-year MACE in B2/C lesions (adjusted HR 0.78; P<.001) but not in A/B1 lesions.
- Benefit was consistent in both NSTEMI (HR 0.73) and STEMI (HR 0.86) for B2/C lesions.
Methodological Strengths
- Large-scale patient-level pooled analysis with stratification by standardized ACC/AHA lesion complexity
- Adjusted hazard models and 3-year follow-up capturing hard endpoints
Limitations
- Nonrandomized observational design with potential selection bias for IVI use
- Registry-based data may have unmeasured confounding and device/procedural heterogeneity
Future Directions: Randomized trials focused on complex AMI lesions to confirm causality, cost-effectiveness analyses, and AI-driven lesion complexity assessment to guide IVI deployment.
INTRODUCTION AND OBJECTIVES: Despite the favorable prognosis associated with intravascular imaging (IVI)-guided percutaneous coronary intervention (PCI) for complex coronary lesions, it is still unclear whether IVI-guided PCI for such lesions provides clinical benefit in patients with acute myocardial infarction (AMI) according to the ACC/AHA lesion classification. METHODS: This study was a patient-level pooled analysis of 2 nationwide Korean AMI registries. We identified 23 051 patients from KAMIR-V and KAMIR-NIH who underwent successful PCI for an infarct-related artery and stratified them by the ACC/AHA lesion classification. Clinical outcomes were compared between IVI-guided and angiography-guided PCI. The primary endpoint was major adverse cardiac events (MACE), a composite of cardiac death, AMI, repeat revascularization, and stent thrombosis, at 3 years. RESULTS: IVI-guided PCI demonstrated a lower incidence of MACE compared with angiography-guided PCI in patients with type B2/C lesions (adjusted HR, 0.78; 95%CI, 0.70-0.88; P < .001), but not in patients with type A/B1 lesions (adjusted HR, 0.81, 95%CI, 0.60-1.11; P = .190). In both non-ST-segment elevation myocardial infarction and ST-segment elevation myocardial infarction, a significantly lower risk of MACE following IVI-guided PCI than angiography-guided PCI was observed in patients with type B2/C lesions (non-ST-segment elevation myocardial infarction: adjusted HR, 0.73; 95%CI, 0.63-0.84; P < .001; ST-segment elevation myocardial infarction: adjusted HR, 0.86, 95%CI, 0.75-0.98; P = .027), but not in those with type A/B1 lesions. CONCLUSIONS: Among patients with AMI, IVI-guided PCI was associated with a significantly lower risk of MACE in those with type B2/C lesions, but not in those with type A/B1 lesions. The prognostic benefit of IVI-guided PCI increased with greater lesion complexity in the infarct-related artery.