Weekly Cardiology Research Analysis
This week’s cardiology literature highlights three actionable advances: a mechanistic preclinical discovery (KIF13B/ITCH/CBL/MERTK axis) that impairs macrophage efferocytosis and is pharmacologically reversible, a methodological breakthrough in multi-ancestry fine-mapping (SuShiE) that sharpens causal variant and gene discovery across populations, and robust clinical trial evidence on procedural strategies and scalable diagnostics (durable hybrid AF ablation and noise‑resilient single‑lead ECG A
Summary
This week’s cardiology literature highlights three actionable advances: a mechanistic preclinical discovery (KIF13B/ITCH/CBL/MERTK axis) that impairs macrophage efferocytosis and is pharmacologically reversible, a methodological breakthrough in multi-ancestry fine-mapping (SuShiE) that sharpens causal variant and gene discovery across populations, and robust clinical trial evidence on procedural strategies and scalable diagnostics (durable hybrid AF ablation and noise‑resilient single‑lead ECG AI / handheld echo AI). Together these reports emphasize translational target discovery, ancestry‑aware genomics, and scalable AI-enabled diagnostics with immediate implications for trial design and care pathways.
Selected Articles
1. The macrophage-derived motor protein KIF13B enhances MERTK-mediated efferocytosis and prevents atherosclerosis in mice.
Human plaque analyses and multiple murine models show that KIF13B deficiency impairs macrophage efferocytosis by lowering ITCH and enabling CBL‑mediated ubiquitination and degradation of MERTK, enlarging plaques independent of plasma lipid changes. Oral CBL antagonism (NX‑1607) restored MERTK, rescued efferocytosis, and reduced plaque burden in vivo, identifying a druggable axis for residual inflammatory risk.
Impact: Uncovers a novel, targetable macrophage pathway that links protein stability (MERTK) to efferocytosis and plaque progression and demonstrates in vivo pharmacologic reversibility — a clear translational path from mechanism to potential therapy.
Clinical Implications: Although preclinical, this identifies CBL antagonism (or strategies to stabilize MERTK) as a candidate approach to reduce plaque inflammation and residual risk beyond lipid lowering; supports advancing to larger animal studies and early‑phase clinical trials focused on efferocytosis endpoints.
Key Findings
- KIF13B expression is reduced in human atherosclerotic plaques and inversely correlates with disease severity in mouse models.
- Myeloid Kif13b deletion enlarged plaques with more macrophage apoptosis and impaired efferocytosis despite unchanged plasma lipids.
- Mechanism: KIF13B loss decreases ITCH expression, allowing CBL-mediated ubiquitination and degradation of MERTK.
- Pharmacologic CBL antagonist NX‑1607 restored MERTK levels, rescued efferocytosis, and reduced atherosclerosis in Kif13b‑deficient mice.
2. Improved multiancestry fine-mapping identifies cis-regulatory variants underlying molecular traits and disease risk.
Introduces SuShiE, a model that leverages linkage disequilibrium heterogeneity across ancestries to improve cis‑molQTL fine‑mapping precision and cross‑ancestry effect estimation. Applied to >36,000 molecular phenotypes, SuShiE fine‑mapped more genes with fewer variants, improved functional enrichment, and increased TWAS/PWAS discovery by ~25% for blood traits, strengthening causal inference for target prioritization.
Impact: Methodological advance that materially improves causal variant/gene resolution across ancestries, directly enabling more reliable target discovery and reducing bias from Eurocentric reference datasets—critical for equitable cardiovascular genomics.
Clinical Implications: Not immediately practice‑changing but accelerates discovery and validation of genetically supported therapeutic targets and biomarkers across ancestries; improves interpretability of TWAS/PWAS used to nominate cardiovascular drug targets.
Key Findings
- SuShiE leverages LD heterogeneity to improve cis‑molQTL fine‑mapping and cross‑ancestry effect inference.
- Fine‑mapping across 36,907 molecular phenotypes mapped 18.2% more genes with fewer prioritized variants and stronger functional enrichment versus existing methods.
- Using SuShiE effect sizes increased TWAS/PWAS gene discovery by ~25.4% for blood cell traits in All of Us.
3. Development and multinational validation of an ensemble deep learning algorithm for detecting and predicting structural heart disease using noisy single-lead electrocardiograms.
ADAPT‑HEART, a noise‑resilient ensemble deep learning model using single‑lead (lead I) ECGs with paired echocardiography, achieved AUROC ≈0.88 in internal testing and generalized to multiple external cohorts (AUROC 0.85–0.89 and 0.859 in ELSA‑Brasil). Among individuals without baseline structural disease, high model probability predicted 2.8–5.7× higher future SHD risk, supporting a role as a predictive screening biomarker deployable on wearables.
Impact: Provides externally validated, wearable‑compatible AI for structural heart disease detection and risk prediction—bridging precision algorithms with population screening and potential to shift referral patterns for echocardiography.
Clinical Implications: Enables low‑friction community and primary care screening to triage high‑risk individuals to echocardiography; prospective implementation studies should assess downstream imaging yield, workflow integration, and clinical outcome impact.
Key Findings
- ADAPT‑HEART achieved AUROC 0.879 for SHD detection in test set with good calibration.
- External validation AUROC ranged 0.852–0.891 across US hospital sites and 0.859 in ELSA‑Brasil.
- High model probability predicted 2.8–5.7‑fold higher incidence of future SHD among those without baseline disease.