Abstract
BACKGROUND: Pulmonary hypertension with left heart failure (PH-LHF) is a prevalent phenotype with distinct hemodynamics. The arrhythmic landscape-encompassing atrial tachyarrhythmias (AT), ventricular tachyarrhythmias (VT), and bradyarrhythmias (BA)-remains poorly characterized. We aimed to identify phenotype-specific risk factors and prognostic impacts of these arrhythmias in PH-LHF. METHODS: This retrospective cohort study included 1,530 PH-LHF patients. Patients were stratified into four mutually exclusive groups based on hierarchical rhythm documentation: No Arrhythmia (NA), AT, VT, and BA. Multivariate Logistic regression and Cox proportional hazards models (adjusting for NT-proBNP, ischemic etiology, and treating device therapy as time-dependent covariates) were utilized to determine independent predictors of each arrhythmia type and to assess their impact on all-cause mortality and heart failure rehospitalization over a median follow-up of 38 months. RESULTS: The overall prevalence of clinically significant arrhythmias was 68.9%, comprising AT (42.2%), VT (18.0%), and BA (8.7%). Left atrial volume index (LAVI) >40 ml/m² (OR 2.41, 95% CI 1.85-3.15) was the strongest predictor of AT. Conversely, a marker of right ventricular dysfunction (TAPSE <16 mm) was a potent independent predictor of VT (OR 3.25, 95% CI 2.38-4.45), largely independent of LVEF. Prognostically, VT was associated with the highest risk of all-cause mortality (HR 2.56, 95% CI 1.95-3.36), whereas AT was the primary driver of rehospitalization (HR 1.92, 95% CI 1.62-2.28). The prevalence of VT increased disproportionately with PH severity. Overlap analysis revealed that 15% of patients in the VT group also had documented AT, and secondary time-dependent analyses confirmed the independent prognostic weight of VT. CONCLUSION: Arrhythmia burden in PH-LHF is substantial and biologically distinct. The "right heart phenotype" drives malignant ventricular arrhythmias and mortality, while the "left atrial phenotype" drives atrial arrhythmias and morbidity. These findings advocate for a precision medicine approach, suggesting that RV monitoring could serve as a valuable tool for risk enrichment in SCD risk stratification in this vulnerable population, pending prospective validation.