Combined assessment of septal scar and septal flash by cardiac magnetic resonance identifies responders to cardiac resynchronization therapy

Abstract

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): South-Eastern Norway Regional Health Authority. Research grants of the University of Leuven. Background Left ventricular (LV) scar, particularly in the lateral wall and septum, reduces response rate to cardiac resynchronization therapy (CRT), whereas a dyssynchronous LV contraction pattern (septal flash) suggests good response. Lateral wall scar abolishes septal flash. Therefore, a combined approach of septal scar and septal flash may characterize the myocardial substrate responsive to CRT. Cardiac magnetic resonance (CMR) may assess both scar and contraction pattern. Purpose The present study aimed to determine if combined assessment of septal scar and septal flash by CMR as single image modality identifies responders to CRT. Methods We investigated all CRT recipients with available CMR from a prospective, multicenter study (n = 136), with both ischemic and non-ischemic heart failure. Septal scar was assessed by late gadolinium enhancement (LGE) from a stack of short axis slices (n = 128) and septal flash determined visually on ordinary cine sequences (n = 136). CRT response was defined as ≥15% reduction in LV end-systolic volume by echocardiography at 6 months follow-up. We also assessed heart transplantation or death of any cause 39 ± 13 months after device implantation. Results In multivariate analysis including percentage septal scar (LGE), septal flash, QRS-duration and QRS-morphology, septal LGE and septal flash were the only independent predictors of CRT response (both p < 0.001). A combined approach of septal LGE and septal flash predicted CRT response with area under the curve 0.86 (95% confidence interval (CI): 0.78-0.94) and long-term survival without heart transplantation with hazard ratio 0.18 (95% CI: 0.05-0.61). A practical approach to selection of CRT candidates by septal LGE and septal flash is illustrated in the present figure. As shown, absence of septal LGE indicated excellent response rate (93%) to CRT independent of other parameters. When septal LGE was present, however, overall response rate was substantially lower (58%), but presence or absence of septal flash separated responders from non-responders with high accuracy. This sequential approach correctly classified 86% of patients. Importantly, the approach was equally accurate in patients with intermediate QRS duration (130-150ms), where 93% of patients were correctly classified. Conclusions Combined assessment of septal LGE and septal flash by CMR as single image modality identifies CRT responders with high accuracy and predicts long-term survival. Abstract Figure.