Daily Cardiology Research Analysis
Three advances span mechanisms, imaging-guided therapy, and population health. A time-resolved transcriptomic study shows circular RNAs rise during vascular cell differentiation and can serve as blood biomarkers for atherosclerotic disease. In complex coronary artery disease, intravascular imaging–guided PCI lowers target vessel failure, with benefits increasing alongside lesion complexity. Street-level deep learning of built environments links greenspace and sidewalks to reduced major adverse c
Summary
Three advances span mechanisms, imaging-guided therapy, and population health. A time-resolved transcriptomic study shows circular RNAs rise during vascular cell differentiation and can serve as blood biomarkers for atherosclerotic disease. In complex coronary artery disease, intravascular imaging–guided PCI lowers target vessel failure, with benefits increasing alongside lesion complexity. Street-level deep learning of built environments links greenspace and sidewalks to reduced major adverse cardiovascular events.
Research Themes
- Mechanistic biomarkers (circular RNAs) for vascular disease
- Intravascular imaging-guided PCI outcomes in complex CAD
- AI analysis of built environment and cardiovascular risk
Selected Articles
1. Circular RNAs increase during vascular cell differentiation and are biomarkers for vascular disease.
Daily RNA-seq across iPSC differentiation showed a global rise in circularization tied to MYC downregulation and reduced splicing factors. In patients, select circRNAs fell in atherosclerotic tissue and PBMCs, and a small panel (COL4A1/2, HSPG2, YPEL2) discriminated disease with AUC 0.79 in tissue and 0.73 in blood.
Impact: Links developmental circRNA dynamics to disease signatures and provides a tangible, minimal circRNA panel with cross-tissue diagnostic potential.
Clinical Implications: CircRNA panels could evolve into noninvasive blood tests for atherosclerosis detection or risk stratification; assays require standardization and prospective validation.
Key Findings
- Global circRNA levels increased during EC and SMC maturation with 397 and 214 circRNAs upregulated >2-fold (adjusted P<0.05).
- MYC decreased during maturation, paralleled by downregulation of splicing factors (e.g., SRSF1, SRSF2) and altered circRNA levels.
- In patients, circRNAs decreased in atherosclerotic tissue and PBMCs; a panel from COL4A1/COL4A2/HSPG2/YPEL2 discriminated disease (tissue AUC 0.79; PBMC AUC 0.73).
Methodological Strengths
- Time-resolved high-throughput RNA-seq with de novo circRNA detection across differentiation trajectory
- External validation in human arterial tissue and PBMCs with machine-learning classification
Limitations
- Clinical validation cohorts and outcomes were limited; no prospective prognostic evaluation
- Causality between MYC/splicing changes and circRNA biogenesis not proven in vivo
Future Directions: Prospective, multi-center validation of circRNA panels, assay standardization, and mechanistic in vivo studies linking circRNA biogenesis to vascular remodeling and outcomes.
2. Outcomes of intravascular imaging-guided percutaneous coronary intervention according to lesion complexity.
Across 4,611 patients from an RCT and registry, IVI-guided PCI reduced 3-year target vessel failure versus angiography guidance regardless of complexity, with HR 0.49 in patients with ≥3 complexity features and HR 0.72 in those with <3. Absolute risk reduction increased with each added complexity feature.
Impact: Strengthens the clinical case for routine IVUS/OCT guidance, especially in highly complex lesions, by quantifying the gradient of benefit.
Clinical Implications: In complex PCI, default IVI guidance should be prioritized; the greater the lesion complexity, the larger the expected event reduction and potential cost-effectiveness.
Key Findings
- Patients with ≥3 complex lesion features had higher TVF risk than those with <3 (11.0% vs 7.2%; HR 1.59, 95% CI 1.28-1.96).
- IVI-guided PCI reduced 3-year TVF versus angiography-guided PCI in both strata (≥3: 7.4% vs 14.4%, HR 0.49; <3: 5.7% vs 8.1%, HR 0.72).
- Absolute risk reduction in TVF increased with accumulating complexity features (interaction p=0.048).
Methodological Strengths
- Large sample combining randomized trial and real-world registry
- Prespecified, hard composite endpoint with 3-year follow-up
Limitations
- Pooled analysis includes non-randomized registry data with potential residual confounding
- Heterogeneity in imaging modalities (IVUS vs OCT) and operator practice not fully controlled
Future Directions: Head-to-head trials stratified by complexity, cost-effectiveness analyses, and modality-specific (IVUS vs OCT) optimization strategies.
3. Deep Learning Analysis of Google Street View to Assess Residential Built Environment and Cardiovascular Risk in a U.S. Midwestern Retrospective Cohort.
In 49,887 adults with 2,083 MACE over 26.86 months, higher visible vertical greenspace (tree-sky index) and sidewalk presence around residences were independently associated with lower MACE risk (HR 0.95 and 0.91, respectively). Deep learning of street-level images enables scalable environmental risk profiling.
Impact: Demonstrates a scalable, low-cost approach to quantify built environments and link them to cardiovascular outcomes, informing public health and urban planning.
Clinical Implications: Incorporate environmental risk context into prevention strategies; clinicians and systems could target high-risk neighborhoods for tailored interventions and advocate for walkability and greenspace.
Key Findings
- Among 49,887 individuals, 2,083 MACE occurred over a median 26.86 months.
- Higher tree-sky index near residences associated with lower MACE risk (HR 0.95, 95% CI 0.91–0.99).
- Sidewalk presence associated with lower MACE risk (HR 0.91, 95% CI 0.87–0.96), independent of greenspace and other covariates.
Methodological Strengths
- Very large cohort with comprehensive adjustment for demographic, socioeconomic, environmental, and clinical covariates
- Novel application of deep learning on street-level imagery to derive exposure metrics (greenspace, sidewalks)
Limitations
- Retrospective, single-region cohort limits causal inference and generalizability
- Potential exposure misclassification from imagery and unmeasured neighborhood confounders
Future Directions: Prospective, multi-city replication; interventional studies on urban design; integration with personal mobility/wearable data to clarify mechanisms.