Erythrocyte Membrane β-Adrenoreactivity as a Systemic Adrenergic Phenotype Associated with Ventricular Tachyarrhythmia in Heart Failure Patients Undergoing Cardiac Resynchronization Therapy

Chronic hyperactivity of the sympathetic nervous system and remodeling of β-adrenergic receptors represent core pathophysiological mechanisms in heart failure with reduced ejection fraction (HFrEF) and in the emergence of ventricular arrhythmias. Nevertheless, clinically applicable functional indicators of this axis are still limited. We investigated whether erythrocyte membrane β-adrenoreactivity (EMA), an indirect functional membrane index derived from β-adrenergic antagonist-modulated hypo-osmotic hemolysis, correlates with a ventricular tachyarrhythmia (VTA) phenotype following cardiac resynchronization therapy with defibrillation capability (CRT-D). In a single-center observational cohort, 99 HFrEF patients underwent CRT-D implantation. A total of 70 patients with complete baseline and 12-month data formed the complete-case analysis (NCT03667989: n = 14; NCT05769036: n = 36; non-registry enrollment under the same institutional protocol: n = 20). Patients were categorized by the prespecified database variable “VTA preCRT and 12 m. post-CRT” into VTA-positive (n = 38) and VTA-negative (n = 32) groups. Baseline EMA was higher in VTA-positive patients (31.4% [22.8; 48.3]) than in VTA-negative patients (22.0% [15.8; 33.4]; p = 0.012). In multivariate logistic regression with stepwise selection, two variables were independently associated with the VTA-positive phenotype: EMA (odds ratio [OR] = 1.0446 per 1%; 95% confidence interval [CI] 1.0092–1.0812; p = 0.013) and left ventricular end-systolic dimension (LVESD) (OR = 1.1023 per 1 mm; 95% CI 1.0102–1.2029; p = 0.029). The model was statistically significant (likelihood-ratio χ2 = 12.567; p = 0.0019) and well calibrated (Hosmer–Lemeshow χ2 = 9.70; p = 0.286), with good discrimination (AUC = 0.748; 95% CI 0.629–0.844). However, explanatory power was modest (Nagelkerke R2 = 0.220), indicating that most variance in the VTA phenotype remains unexplained by EMA and LVESD. At a cut-off of 0.5, sensitivity was 73.68%, specificity 71.87%, and overall accuracy 72.86%. Internal validation with 2000 bootstrap iterations confirmed robustness (optimism-corrected AUC = 0.727; 95% CI 0.612–0.835). EMA remained significant after adjustment for LVESD, supporting its role as an independent marker of the VTA-positive phenotype. The model outperformed all univariate models in terms of AUC, R2, and overall accuracy. EMA may therefore be associated with a systemic adrenergic membrane phenotype and an arrhythmic substrate. These findings require mechanistic validation and external replication before any clinical application. All reported cut-offs are exploratory and derived from a single center, and the analysis is based on a complete-case cohort with acknowledged potential for attrition bias.