Stratifying heart failure patients with graph neural network and transformer using Electronic Health Records to optimize drug response prediction.

OBJECTIVES: Heart failure (HF) impacts millions of patients worldwide, yet the variability in treatment responses remains a major challenge for healthcare professionals. The current treatment strategies, largely derived from population based evidence, often fail to consider the unique characteristics of individual patients, resulting in suboptimal outcomes. This study aims to develop computational models that are patient-specific in predicting treatment outcomes, by utilizing a large Electronic Health Records (EHR) database. The goal is to improve drug response predictions by identifying specific HF patient subgroups that are likely to benefit from existing HF medications. MATERIALS AND METHODS: A novel, graph-based model capable of predicting treatment responses, combining Graph Neural Network and Transformer was developed. This method differs from conventional approaches by transforming a patient's EHR data into a graph structure. By defining patient subgroups based on this representation via K-Means Clustering, we were able to enhance the performance of drug response predictions. RESULTS: Leveraging EHR data from 11 627 Mayo Clinic HF patients, our model significantly outperformed traditional models in predicting drug response using NT-proBNP as a HF biomarker across five medication categories (best RMSE of 0.0043). Four distinct patient subgroups were identified with differential characteristics and outcomes, demonstrating superior predictive capabilities over existing HF subtypes (best mean RMSE of 0.0032). DISCUSSION: These results highlight the power of graph-based modeling of EHR in improving HF treatment strategies. The stratification of patients sheds light on particular patient segments that could benefit more significantly from tailored response predictions. CONCLUSIONS: Longitudinal EHR data have the potential to enhance personalized prognostic predictions through the application of graph-based AI techniques.

Successful prediction of left bundle branch block-induced cardiomyopathy and treatment effect by artificial intelligence-enabled electrocardiogram.

BACKGROUND: Left bundle branch block (LBBB) induced cardiomyopathy is an increasingly recognized disease entity.  However, no clinical testing has been shown to be able to predict such an occurrence. CASE REPORT: A 70-year-old male with a prior history of LBBB with preserved ejection fraction (EF) and no other known cardiovascular conditions presented with presyncope, high-grade AV block, and heart failure with reduced EF (36%). His coronary angiogram was negative for any obstructive disease. No other known etiologies for cardiomyopathy were identified. Artificial intelligence-enabled ECGs performed 6 years prior to clinical presentation consistently predicted a high probability (up to 91%) of low EF. The patient successfully underwent left bundle branch area (LBBA) pacing with correction of the underlying LBBB. Subsequent AI ECGs showed a large drop in the probability of low EF immediately after LBBA pacing to 47% and then to 3% 2 months post procedure. His heart failure symptoms markedly improved and EF normalized to 54% at the same time. CONCLUSIONS: Artificial intelligence-enabled ECGS may help identify patients who are at risk of developing LBBB-induced cardiomyopathy and predict the response to LBBA pacing.

Left bundle branch pacing vs ventricular septal pacing for cardiac resynchronization therapy.

Background: The outcomes of left bundle branch pacing (LBBP) and left ventricular septal pacing (LVSP) in patients with heart failure remain to be learned. Objective: The objective of this study was to assess the echocardiographic and clinical outcomes of LBBP, LVSP, and deep septal pacing (DSP). Methods: This retrospective study included patients who met the criteria for cardiac resynchronization therapy (CRT) and underwent attempted LBBP in 5 Mayo centers. Clinical, electrocardiographic, and echocardiographic data were collected at baseline and follow-up. Results: A total of 91 consecutive patients were included in the study. A total of 52 patients had LBBP, 25 had LVSP, and 14 had DSP. The median follow-up duration was 307 (interquartile range 208, 508) days. There was significant left ventricular ejection fraction (LVEF) improvement in the LBBP and LVSP groups (from 35.9 ± 8.5% to 46.9 ± 10.0%, P < .001 in the LBBP group; from 33.1 ± 7.5% to 41.8 ± 10.8%, P < .001 in the LVSP group) but not in the DSP group. A unipolar paced right bundle branch block morphology during the procedure in lead V1 was associated with higher odds of CRT response. There was no significant difference in heart failure hospitalization and all-cause deaths between the LBBP and LVSP groups. The rate of heart failure hospitalization and all-cause deaths were increased in the DSP group compared with the LBBP group (hazard ratio 5.10, 95% confidence interval 1.14-22.78, P = .033; and hazard ratio 7.83, 95% confidence interval 1.38-44.32, P = .020, respectively). Conclusion: In patients undergoing CRT, LVSP had comparable CRT outcomes compared with LBBP.

Kinetically Controlled Synthesis of Quasi-Square CsPbI3 Nanoplatelets with Excellent Stability.

Nanoplatelets (NPLs) share excellent luminescent properties with their symmetric quantum dots counterparts and entail special characters benefiting from the shape, like the thickness-dependent bandgap and anisotropic luminescence. However, perovskite NPLs, especially those based on iodide, suffer from poor spectral and phase stability. Here, stable CsPbI3 NPLs obtained by accelerating the crystallization process in ambient-condition synthesis are reported. By this kinetic control, the rectangular NPLs into quasi-square NPLs are tuned, where enlarged width endows the NPLs with a lower surface-area-to-volume ratio (S/V ratio), leading to lower surficial energy and thus improved endurance against NPL fusion (cause for spectral shift or phase transformation). The accelerated crystallization, denoting the fast nucleation and short period of growth in this report, is enabled by preparing a precursor with complete transformation of PbI2 into intermediates (PbI3 -), through an additional iodide supplier (e.g., zinc iodide). The excellent color stability of the materials remains in the light-emitting diodes under various bias stresses.

Feasibility and Safety Study of Concomitant Left Bundle Branch Area Pacing and Atrioventricular Node Ablation with Same-Day Hospital Dismissal.

BACKGROUND: Left bundle branch area pacing (LBBAP) has rapidly emerged as a promising modality of physiologic pacing and has demonstrated excellent lead stability. In this retrospective study, we evaluate whether this pacing modality can allow concomitant atrioventricular node (AVN) ablation and same-day dismissal. METHODS: Twenty-four consecutive patients (female 63%, male 37%) with an average age of 78 ± 5 years were admitted for pacemaker (75%)/defibrillator (25%) implantations and concomitant AVN ablation. Device implantation with LBBAP was performed first, followed by concomitant AVN ablation through left axillary vein access to allow for quicker post-procedure ambulation. The patients were discharged on the same day after satisfactory post-ambulation device checks. RESULTS: LBBAP was successful in 22 patients (92% in total, 20 patients had an LBBP and two patients had a likely LBBP), followed by AVN ablation from left axillary vein access (21/24, 88%). All patients had successful post-op chest x-rays, post-ambulation device checks, and were discharged on the same day. After a mean follow up of three months, no major complications occurred, such as LBBA lead dislodgement requiring a lead revision. The LBBA lead pacing parameters immediately after implantation vs. three-month follow up were a capture threshold of 0.8 ± 0.3 V@0.4 ms vs. 0.6 ± 0.3 V@0.4 ms, sensing 9.9 ± 3.9 mV vs. 10.4 ± 4.1 mV, and impedance of 710 ± 216 ohm vs. 544 ± 110 ohm. The QRS duration before and after AVN ablation was 117 ± 32 ms vs. 123 ± 14 ms. Mean LVEF before and three months after the implantation was 44 ± 14% vs. 46 ± 12%. CONCLUSION: LBBA pacing not only offers physiologic pacing, but also allows for a concomitant AVN ablation approach from the left axillary vein and safe same-day hospital dismissal.

Chinese herbal medicine and active ingredients for diabetic cardiomyopathy: molecular mechanisms regulating endoplasmic reticulum stress.

Background: Diabetic cardiomyopathy (DCM) is one of the serious microvascular complications of diabetes mellitus. It is often associated with clinical manifestations such as arrhythmias and heart failure, and significantly reduces the quality of life and years of survival of patients. Endoplasmic reticulum stress (ERS) is the removal of unfolded and misfolded proteins and is an important mechanism for the maintenance of cellular homeostasis. ERS plays an important role in the pathogenesis of DCM by causing cardiomyocyte apoptosis, insulin resistance, calcium imbalance, myocardial hypertrophy and fibrosis. Targeting ERS is a new direction in the treatment of DCM. A large number of studies have shown that Chinese herbal medicine and active ingredients can significantly improve the clinical outcome of DCM patients through intervention in ERS and effects on myocardial structure and function, which has become one of the hot research directions. Purpose: The aim of this review is to elucidate and summarize the roles and mechanisms of Chinese herbal medicine and active ingredients that have the potential to modulate endoplasmic reticulum stress, thereby contributing to better management of DCM. Methods: Databases such as PubMed, Web of Science, China National Knowledge Internet, and Wanfang Data Knowledge Service Platform were used to search, analyze, and collect literature, in order to review the mechanisms by which phytochemicals inhibit the progression of DCM by targeting the ERS and its key signaling pathways. Keywords used included "diabetic cardiomyopathy" and "endoplasmic reticulum stress." Results: This review found that Chinese herbs and their active ingredients can regulate ERS through IRE1, ATF6, and PERK pathways to reduce cardiomyocyte apoptosis, ameliorate myocardial fibrosis, and attenuate myocardial hypertrophy for the treatment of DCM. Conclusion: A comprehensive source of information on potential ERS inhibitors is provided in this review. The analysis of the literature suggests that Chinese herbal medicine and its active ingredients can be used as potential drug candidates for the treatment of DCM. In short, we cannot ignore the role of traditional Chinese medicine in regulating ERS and treating DCM, and look forward to more research and new drugs to come.

Left bundle branch area pacing too far away from the tricuspid annulus may cause pacing induced cardiomyopathy.

BACKGROUND: Physiologic pacing through left bundle branch area pacing (LBBAP) has recently been shown to be a very promising alternative for cardiac resynchronization therapy (CRT) and to avoid pacing induced cardiomyopathy. However, it is not clear whether the position of LABBP lead may affect the clinical outcomes. CASE REPORT: We here report a case of likely LBBAP induced worsening heart failure and cardiomyopathy reversed by re-positioning of the pacing lead towards a more annular position. A 70-year-old male with a previous history of non-ischemic dilated cardiomyopathy (ejection fraction 40%) who developed intermittent complete heart block and required permanent ventricular pacing. LBBAP was performed with the lead positioned to a position relatively far away from the tricuspid annulus (3.7 cm), due to difficulty in fixating the lead deep into the septum at a more annular position. One month post procedure, the patient's heart failure symptoms worsened, and his EF decreased to 31% despite good heart failure management. He underwent CRT upgrade with successful revision of the originally implanted LBBAP lead to a more annular position, using a deflectable delivery sheath. This resulted in further narrowing of the paced QRS duration from 135 to 106 ms. Two months post procedure, his heart failure symptoms improved by one functional class, and EF improved to 41% by echocardiogram. CONCLUSIONS: LBBAP may be harmful when the lead is placed too far away from the annulus and may cause paced induced cardiomyopathy.