Towards an optimal mechanical activation delay for left ventricular lead placement in cardiac resynchronisation therapy

Background/introduction Left ventricular (LV) lead position is a key determinant of response to cardiac resynchronisation therapy (CRT). Whilst targeting the latest mechanically activated region has been advocated, it has also been suggested that pacing a site that yields the most rapid and homogeneous LV activation may be more effective. Purpose Using cardiac magnetic resonance (CMR)-based measurements, we investigated the effect of LV lead placement in a segment with earlier mechanical activation than the latest activated segment on left ventricular end-systolic volume (LVESV) reduction. Methods We included 103 patients with a class I or IIa indication for CRT who had cine CMR of sufficient quality, as well as five healthy subjects without an indication for CRT. Radial strain analysis was performed to determine mechanical activation delay in a 36-segment cardiac model, excluding the septal segments. For all subjects, time to peak strain (TTP) was determined and divided by aortic valve closure (AVC) time to derive the segmental mechanical dyssynchrony (MD) index (Figure 1). An MD index greater than 1.0 indicates peak mechanical contraction after AVC. The MD index was plotted on the 36-segment cardiac model. For CRT patients, the MD index of the segment in which the LV electrode was implanted was correlated with LVESV reduction (%) after six months of CRT. Linear regression of the absolute distance from the MD index was applied to identify the MD index at which the model best predicted LVESV reduction (highest adjusted R² with a negative slope). Results Healthy controls showed a narrow MD distribution (1.04 ± 0.07), whereas CRT patients exhibited a wider range of values at the LV lead segment (mean 1.15 ± 0.25), indicating delayed activation and greater mechanical dispersion. Linear regression identified an optimal MD index of 1.17 for LV lead implantation. LVESV reduction improved with increasing MD up to 1.17, after which it declined. Greater deviation from this value was significantly associated with worse LVESV reduction (β = -31.8, 95% CI -54.1 to -9.5, p = 0.006). When the MD index of the segment containing the LV lead was divided into three groups (<1.10 as too early, 1.10-1.24 as optimally, and >1.24 as too late), median LVESV reduction was significantly greater (p = 0.046) in the optimally group compared with the too late group (Figure 1). Figure 2 illustrates a responder with an LV lead implanted in a segment with an MD index of 1.17 (optimally) and a non-responder with the LV lead in a segment with an MD index of 2.24 (too late). Conclusion Our analysis demonstrates that there is an optimal degree of dyssynchrony which maximises CRT response, beyond which additional delay is associated with diminished remodelling. These findings suggest that the MD index may be useful for targeted LV lead implantation to optimise CRT response.