The value of serum metabolomics analysis in predicting the response to cardiac resynchronization therapy

OBJECTIVE: To construct a prediction model based on metabolic profiling for predicting the response to cardiac resynchronization therapy (CRT). METHODS: Peripheral venous (PV) and coronary sinus (CS) blood samples were collected from 25 patients with heart failure (HF) at the time of CRT implantation, and PV blood samples were obtained from ten healthy controls. The serum samples were analyzed by liquid chromatography-mass spectrometry (LC-MS). As per the clinical and echocardiographic assessment at the 6-month follow-up, the HF patients were categorized as CRT responders and non-responders. RESULTS: HF patients had altered serum metabolomic profiles that were significantly different from those of the healthy controls. Differential metabolites were also observed between CRT responders and non-responders. A prediction model for CRT response (CRT-Re) was constructed using the concentration levels of the differential metabolites, L-arginine and taurine. The optimal cutoff value of the CRT-Re model was found to be 0.343 by ROC analysis (sensitivity, 88.2%; specificity, 87.5%; Area under curve (AUC) = 0.897, P = 0.002). The concentration levels of the differential metabolites, L-arginine and lysyl-gamma-glutamate, in PV serum were significantly correlated with that in CS serum (r = 0.945 and 0.680, respectively, all P < 0.001). CONCLUSIONS: Our results suggest that serum-based metabolic profiling may be a potential complementary screening tool for predicting the outcome of CRT.