Daily Cardiology Research Analysis
Analyzed 158 papers and selected 3 impactful papers.
Summary
Three papers stood out today: a prospective proteomics study shows that an Lp(a)-associated plasma protein signature predicts atherosclerotic events beyond Lp(a) concentration; a fully automated deep learning system quantifies coronary plaque on CCTA with strong external validation and prognostic value; and a prospective multicenter registry reports substantial reduction in ventricular tachycardia burden after stereotactic arrhythmia radioablation with an acceptable safety profile.
Research Themes
- Proteomics-based risk stratification for ASCVD
- Automated coronary plaque quantification and prognosis
- Noninvasive stereotactic radiotherapy for refractory ventricular tachycardia
Selected Articles
1. Lipoprotein(a)-Associated Proteomic Signature Predicts Cardiovascular Disease in Young Adults.
In 3,920 CARDIA participants with 27-year follow-up, an Lp(a)-associated proteomic signature reflecting immune activation, coagulation, and vascular dysfunction predicted coronary artery calcification and incident CHD independent of Lp(a) concentration. The score replicated in 37,996 UK Biobank participants and was associated with CRP, incident CHD, and all-cause mortality, suggesting added biological and prognostic information beyond Lp(a) levels.
Impact: This study advances precision prevention by integrating proteomics with a lifelong cohort and external replication to refine ASCVD risk prediction in young adults beyond a single genetic biomarker.
Clinical Implications: Proteomic signatures may augment risk stratification for early ASCVD, informing selection and timing of Lp(a)-lowering and other preventive therapies in young adults with elevated Lp(a).
Key Findings
- Lp(a) was associated with CAC (OR 1.23 [1.13–1.34]) and incident CHD (HR 1.23 [1.07–1.41]).
- An Lp(a)-associated proteomic score predicted CAC (standardized beta 0.40, p<0.0001) and hs-CRP (beta 0.11, p=0.00015) independent of Lp(a).
- In UK Biobank, a recalibrated proteomic score associated with CRP, incident CHD, and all-cause mortality, replicating prognostic utility.
Methodological Strengths
- Prospective cohort with 27-year follow-up and prespecified outcomes (CAC, incident CHD).
- External replication in a large independent cohort (UK Biobank, n=37,996) and multivariable modeling with LASSO.
Limitations
- Observational design cannot establish causality of proteomic pathways.
- Proteomic platform and assay variability may limit generalizability; clinical thresholds for implementation are not defined.
Future Directions: Validate the proteomic score across ancestries and clinical settings, integrate with imaging and genetics, and test whether proteomic-guided prevention improves outcomes.
BACKGROUND: Elevated lipoprotein(a) [Lp(a)] is associated with a higher risk of atherosclerotic cardiovascular disease (ASCVD). Although Lp(a) is a genetically determined risk factor, the plasma proteomic features associated with Lp(a) and whether they provide information about ASCVD risk beyond Lp(a) concentration are not well characterized. OBJECTIVE: We sought to identify plasma proteomic features associated with Lp(a) concentration and to evaluate whether an Lp(a)-associated proteomic signature is associated with ASCVD phenotypes in young, healthy adults. METHODS: In the Corona
2. Stereotactic arrhythmia radioablation for refractory ventricular tachycardia: the STOPSTORM.eu study.
In 193 patients across 28 centers (median follow-up 19 months), STAR reduced sustained VT burden by 80% at 6 months among 107 evaluable patients and 72% of survivors ≥6 months were free of ICD shocks. Only 12 serious adverse events were possibly or probably related to treatment, supporting the safety of this noninvasive adjunct in refractory VT.
Impact: This is the largest prospective multicenter evaluation of STAR, demonstrating substantial VT burden reduction with acceptable safety, informing adoption and trial design for noninvasive ablation in refractory VT.
Clinical Implications: STAR may be considered in selected patients with drug- and catheter-ablation–refractory VT to reduce arrhythmia burden and ICD shocks, particularly where invasive options are limited or high risk.
Key Findings
- Median VT episode burden was reduced by 80% comparing 6 months pre- vs post-STAR among evaluable patients.
- Among patients surviving ≥6 months, 72% were free from ICD shocks.
- Only 12 serious adverse events were adjudicated as possibly or probably treatment-related across the full cohort.
Methodological Strengths
- Prospective, international multicenter registry with predefined endpoints.
- Clinically meaningful outcomes including VT burden and ICD shocks; independent SAE adjudication.
Limitations
- Nonrandomized design without a control group limits causal inference.
- Potential selection and center-effects; long-term durability beyond median 19 months requires further follow-up.
Future Directions: Randomized comparative trials versus catheter ablation or optimized medical therapy, standardized STAR workflows, and long-term safety surveillance (coronary, pericardial, arrhythmogenic effects).
BACKGROUND AND AIMS: Stereotactic arrhythmia radioablation (STAR) is increasingly used for refractory ventricular tachycardia (VT), yet prospective multicentre outcome data remain limited. Here, the planned interim analysis of the prospective Standardized Treatment and Outcome Platform for Stereotactic Therapy Of Re-entrant tachycardia by a Multidisciplinary (STOPSTORM) registry is reported. METHODS: STOPSTORM is a European prospective, international, multicentre registry of patients treated with STAR. The primary efficacy endpoint was the change in sustained VT episode burden co
3. A Fully Automated Deep Learning Model for Quantifying Coronary Plaque at Coronary CT Angiography.
PlaqueSegNet achieved excellent agreement with IVUS and expert readers across four external datasets (ICC >0.90), and despite wide Bland-Altman limits, demonstrated prognostic value for MACE (C-index 0.64–0.74) across three independent cohorts with median follow-up 2.3–5.3 years. The fully automated pipeline enables scalable plaque burden quantification in routine CCTA.
Impact: This work tackles a key translation gap by delivering a fully automated, externally validated plaque quantification tool with demonstrated prognostic utility across multiple cohorts and imaging platforms.
Clinical Implications: Automated CCTA plaque quantification could standardize plaque burden assessment, support risk stratification, and enable longitudinal therapy monitoring without additional testing.
Key Findings
- Excellent agreement with IVUS and expert readers across four external datasets (all ICC >0.90).
- Prognostic performance for MACE with C-index 0.64 (China CT-FFR 2), 0.65 (China CT-FFR 1.1), and 0.74 (serial CCTA cohort).
- Robustness across scanner types, serial scans, and photon-counting CT, despite wide Bland–Altman limits of agreement.
Methodological Strengths
- Large multi-institutional development with four independent external test datasets including IVUS ground truth.
- Prognostic validation across three cohorts with multi-year follow-up using Harrell C-index.
Limitations
- Bland–Altman limits of agreement were wide, indicating measurement variability at individual level.
- Retrospective design for model development; prospective clinical impact and workflow integration were not tested.
Future Directions: Prospective, randomized or pragmatic studies to test risk-guided management using automated plaque burden, and integration with CT-FFR and clinical risk scores.
Background Deep learning (DL) models for quantifying plaques at coronary CT angiography (CCTA) are rarely used in routine clinical care. Purpose To develop a fully automated DL model for coronary plaque quantification and to evaluate its prognostic value. Materials and Methods Patients who underwent CCTA were retrospectively enrolled from 17 Chinese hospitals between June 2009 and May 2024. The imaging data of these patients were randomly split into training and validation sets at a 7:3 ratio to develop a fully automated DL model for quantifying plaque volume (PV), PlaqueSegNet, which was subsequent