Myocardial single photon emission computed tomography imaging to identify non-responders to cardiac resynchronization therapy in patients with non-left bundle branch block

Abstract Background Non-response to cardiac resynchronization therapy (CRT) is an important issue in the treatment of heart failure with reduced ejection fraction (HFrEF) and non-left bundle branch block (LBBB). Although the QRS duration and morphology are crucial parameters for selecting CRT candidates, electrocardiogram (ECG) markers may inadequately represent the mechanical activation patterns within the heart. Owing to limitations associated with the ECG criteria, there is increasing interest in advanced multimodality imaging to improve patient selection. Recently, myocardial single-photon emission computed tomography imaging (G-MPI) is used for assessing left ventricular (LV) dyssynchrony. This technique is mostly automatic and has high intra- and inter-observer reproducibility and repeatability. Objectives This study aimed to examine whether preoperative G-MPI evaluation could predict non-responsiveness to CRT in patients with HFrEF with non-LBBB. Methods Between January 2021 and December 2022, 128 patients underwent CRT, of whom 73 with pre-operative evaluation using G-MPI were selected. Forty-three patients with non-LBBB (21 and 22 CRT responders and non-responders, respectively) were assessed. These patients were divided into CRT responder and non-responder groups. We evaluated the cut-off and clinical impact of pre-operative G-MPI parameters in predicting non-responsiveness to CRT. In addition, 30 patients with LBBB were analyzed. Results Among patients with non-LBBB, CRT responders and non-responders showed no significant differences in baseline characteristics, except for the LV dimension. A receiver operating characteristic curve analysis identified 108° and 27.7° as the optimal cut-offs for the bandwidth and phase standard deviation (SD), respectively, to predict non-responsiveness to CRT (area under the curve [AUC]=0.762; 95% confidence interval [CI]: 0.601–0.923 and AUC=0.742; 95% CI: 0.576–0.909, respectively). A multivariate analysis revealed that a cut-off bandwidth of ≥108° and phase SD of ≥27.7° are independent predictors of non-responsiveness to CRT in patients with non-LBBB (hazard ratios 5.65; 95% CI: 1.53–20.9; P=0.009). In contrast, there were no relationships between G-MPI parameters and non-responsiveness to CRT in patients with LBBB. Conclusions Pre-operative G-MPI findings were associated with non-responsiveness to CRT in patients with non-LBBB. The identification of potentially nonresponsive patients could influence patient management and alter outcomes.