Daily Cardiology Research Analysis
Analyzed 68 papers and selected 3 impactful papers.
Summary
Three impactful cardiology studies span therapy, risk stratification, and AI-driven prognostication. A meta-analysis shows advanced extravalvular cardiac damage markedly increases mortality after TAVI. A prespecified EMPULSE analysis supports in-hospital initiation of empagliflozin across de novo and acutely decompensated heart failure. An AI model integrating 3D spatiotemporal CMR with clinical data accurately predicts 5-year MACE after AMI.
Research Themes
- Extravalvular cardiac damage staging to refine TAVI risk stratification
- Early initiation of SGLT2 inhibitors during hospitalization for acute heart failure
- Multimodal AI using 3D spatiotemporal CMR to predict long-term MACE post-AMI
Selected Articles
1. Impact of advanced cardiac damage in severe aortic stenosis on short-term and mid-term mortality and rehospitalisation after transcatheter aortic valve implantation: a systematic review and meta-analysis.
Across 34 studies (n=26,076), extravalvular cardiac damage—particularly right ventricular dysfunction (stage 4)—was associated with substantially higher 12-month all-cause and cardiovascular mortality after TAVI. Even borderline-stage damage conferred excess risk, supporting routine incorporation of cardiac damage staging into preprocedural risk assessment.
Impact: This meta-analysis quantifies the prognostic gradient of extravalvular cardiac damage after TAVI, providing actionable evidence to refine patient selection, counseling, and follow-up intensity.
Clinical Implications: Preprocedural assessment should systematically stage extravalvular damage (e.g., RVD, PH, TR) to inform TAVI risk discussions and tailor surveillance; stage 4 patients may require intensified management and multidisciplinary planning.
Key Findings
- Twelve-month all-cause mortality rose with cardiac damage severity: HR 1.61 (borderline), 2.06 (stage 3), and 2.77 (stage 4).
- Cardiovascular mortality was highest in stage 4 (HR 3.13; RR 2.63).
- Rehospitalisation risk increased in stage 3 (RR 1.33), though data were limited.
- Findings were consistent across valve types and diagnostic modalities; heterogeneity partly explained by age, sex, and comorbidities.
Methodological Strengths
- Prospective registration (PROSPERO) and risk-of-bias assessment with Joanna Briggs Institute checklist
- Random-effects meta-analysis with subgroup and meta-regression analyses across 26,076 patients
Limitations
- Predominantly observational cohorts may introduce residual confounding
- Heterogeneity in definitions of cardiac damage stages and limited rehospitalisation data
Future Directions: Prospective studies should validate standardized cardiac damage staging and evaluate tailored peri-TAVI strategies (e.g., RV optimization) to mitigate excess risk.
BACKGROUND: Severe aortic stenosis (AS) is commonly associated with advanced cardiac damage, including right ventricular dysfunction (RVD), pulmonary hypertension (PH) and tricuspid regurgitation (TR), which may worsen prognosis after transcatheter aortic valve implantation (TAVI). This systematic review and meta-analysis aimed to assess the effect of these conditions on short-term and mid-term mortality and rehospitalisation following TAVI. METHODS: We conducted a systematic search of PubMed, Scopus and Web of Science for studies published up to June 2025. Eligible studies included adults with AS undergoing TAVI and reported outcomes at 1 month, 6 months or 12 months stratified by the presence of RVD, PH or TR. Studies had to report either HRs, risk ratios (RRs) or sufficient raw event data for mortality or rehospitalisation. Data were synthesised using a random-effects meta-analysis. Subgroup analyses were conducted by cardiac damage severity according to the Généreux staging system and stratified by valve type and diagnostic modality. Risk of bias in included studies was assessed using the Joanna Briggs Institute's checklist for cohort studies. Meta-regression was performed to explore sources of between-study heterogeneity. RESULTS: A total of 34 studies including 26 076 patients met inclusion criteria. Twelve-month HRs for all-cause mortality increased with advancing cardiac damage: borderline stage HR 1.61 (1.22-2.12), stage 3 HR 2.06 (1.63-2.60) and stage 4 HR 2.77 (2.11-3.64). RRs followed a similar trend. Cardiovascular mortality was highest in stage 4 (HR 3.13 (1.20-8.17); RR 2.63 (1.54-4.47)). Rehospitalisation data were limited but suggested elevated risk in stage 3 (RR 1.33 (1.12-1.58)). Meta-regression indicated that age, sex and comorbidities contributed to between-study heterogeneity, particularly in stage 3 analyses. CONCLUSION: Extravalvular cardiac damage, especially RVD (stage 4), is strongly associated with increased short-term and mid-term mortality and rehospitalisation after TAVI. Even borderline-stage patients face elevated risk, underscoring the continuous nature of AS-related cardiac injury. Incorporating cardiac damage staging into preprocedural assessment can enhance risk stratification and guide management to improve patient outcomes. PROSPERO REGISTRATION NUMBER: CRD420250638838.
2. 3D Spatiotemporal cardiac reconstruction for predicting MACE in acute myocardial infarction.
Using a novel pipeline that reconstructs temporally-resolved 3D bi-ventricular cine CMR (ReconSeg3D) and integrates imaging with 45 clinical/CMR variables via cross-attention, HeartTTable achieved a 5-year AUC of 0.934 and C-index of 0.897 for MACE after AMI-PCI. The model outperformed tabular-only approaches and enabled robust postoperative risk stratification.
Impact: This work operationalizes advanced spatiotemporal CMR information for long-term risk prediction, demonstrating clear performance gains from true multimodal integration in post-AMI care.
Clinical Implications: If externally validated and deployed, the model could support individualized 5-year risk stratification after AMI-PCI, guiding follow-up intensity and secondary prevention strategies.
Key Findings
- ReconSeg3D reconstructed temporally-resolved 3D bi-ventricular cine CMR sequences from short-axis stacks.
- HeartTTable integrated 3D CMR dynamics with 45 clinical/CMR variables via spatiotemporal decomposition and cross-attention.
- Achieved 5-year time-dependent AUC 0.934 (95% CI 0.907-0.959) and Harrell's C-index 0.897, outperforming tabular-only baselines.
- Demonstrated strong postoperative risk stratification for AMI patients.
Methodological Strengths
- True multimodal fusion of dynamic 3D CMR with clinical variables using cross-attention
- Long-term prognostication with time-dependent AUC and C-index reporting
Limitations
- External validation and multi-center generalizability were not detailed in the abstract
- CMR availability and standardization may limit broad implementation; risk of overfitting needs assessment
Future Directions: Conduct multi-center external validation, assess calibration and decision-curve utility, and evaluate clinical impact via prospective implementation studies.
Artificial intelligence has made significant strides in predicting major adverse cardiovascular events (MACE) in patients with acute myocardial infarction (AMI) following percutaneous coronary intervention. However, most existing methods rely solely on tabular variables derived from clinical data and cardiac magnetic resonance (CMR), without fully leveraging the predictive potential of the CMR imaging modality itself. Moreover, these approaches often overlook the synergistic benefits of multimodal integration between imaging and tabular data. In addition, current models primarily focus on short-term MACE risk assessment (e.g., within 6 months or 1 year), limiting their applicability for long-term prognostication. To address these limitations, we first developed ReconSeg3D, a model that reconstructs short-axis cine CMR stacks into temporally-resolved 3D bi-ventricular volumes, capturing fine-grained cardiac anatomy and dynamic motion. These bi-ventricular sequences were then integrated with 45 clinical and CMR-derived variables using spatiotemporal decomposition and cross-attention mechanisms to construct a multimodal MACE prediction model-HeartTTable. HeartTTable achieved a 5-year time-dependent AUC of 0.934 (95% CI 0.907-0.959) and a Harrell's C-index of 0.897 for predicting MACE risk, significantly outperforming models based solely on clinical and CMR-derived tabular features, and demonstrated strong capabilities in postoperative risk stratification. Our study contributes to improved long-term postoperative management for AMI patients by offering clinicians an objective, data-driven decision-support tool.
3. Empagliflozin in De Novo vs Acute Decompensated Chronic Heart Failure: A Prespecified Analysis From EMPULSE.
In this prespecified EMPULSE analysis, empagliflozin initiated in-hospital yielded favorable hierarchical outcomes at 90 days in both de novo and acutely decompensated chronic HF subgroups. Benefits were similar despite a reduced diuretic response in ADHF, and treatment was well tolerated.
Impact: It supports a generalized practice of initiating SGLT2 inhibitors during hospitalization for acute HF irrespective of de novo versus decompensated status.
Clinical Implications: Hospitals should consider routine in-hospital initiation of empagliflozin in stabilized acute HF to improve short-term clinical status and outcomes, with monitoring for tolerability.
Key Findings
- Randomized, stratified comparison showed similar benefits of empagliflozin in NHF and ADHF at 90 days.
- Win ratio favored empagliflozin: 1.29 (95% CI 0.89-1.89) in NHF and 1.39 (95% CI 1.07-1.81) in ADHF.
- ADHF patients had reduced diuretic response, yet clinical benefit was preserved; tolerability was good.
Methodological Strengths
- Prespecified subgroup analysis within a randomized, placebo-controlled trial
- Hierarchical composite endpoint with win ratio, stratified by HF status
Limitations
- Subgroup analysis not powered for definitive interaction testing
- Follow-up limited to 90 days; NHF win ratio CI included 1
Future Directions: Assess longer-term outcomes and diuretic response dynamics; evaluate implementation strategies for universal in-hospital SGLT2 initiation in acute HF.
BACKGROUND: In EMPULSE (A Study to Test the Effect of Empagliflozin in Patients Who Are in Hospital for Acute Heart Failure), the sodium-glucose cotransporter 2 inhibitor empagliflozin improved clinical outcomes in patients hospitalized for heart failure (HF). OBJECTIVES: This prespecified analysis examined efficacy, safety, and tolerability of empagliflozin in subgroups with de novo heart failure (NHF) vs acute decompensated heart failure (ADHF). METHODS: After stabilization, participants were randomized 1:1 to empagliflozin 10 mg/d or placebo, stratified by HF status (NHF: n = 175; ADHF: n = 355). The primary endpoint was a hierarchical composite of death, worsening HF, or ≥5-point difference in Kansas City Cardiomyopathy Questionnaire-Total Symptom Score (KCCQ-TSS) change at day 90, assessed using a win ratio. RESULTS: Participants with NHF were younger, had fewer comorbidities, had higher blood pressure and heart rate, and better KCCQ-TSS. Prescription of diuretic agents was similar between subgroups. The win ratio was 1.29 (95% CI: 0.89-1.89) for NHF and 1.39 (95% CI: 1.07-1.81) for ADHF (P CONCLUSIONS: In-hospital initiation of empagliflozin produced similar clinical benefits in NHF and ADHF despite the reduced diuretic response in participants with ADHF and was well tolerated. This supports in-hospital initiation of empagliflozin in all patients with acute HF (A Study to Test the Effect of Empagliflozin in Patients Who Are in Hospital for Acute Heart Failure [EMPULSE]; NCT04157751).