Electrical latency predicts the optimal left ventricular endocardial pacing site: results from a multicentre international registry.

Aims: The optimal site for biventricular endocardial (BIVENDO) pacing remains undefined. Acute haemodynamic response (AHR) is reproducible marker of left ventricular (LV) contractility, best expressed as the change in the maximum rate of LV pressure (LV-dp/dtmax), from a baseline state. We examined the relationship between factors known to impact LV contractility, whilst delivering BIVENDO pacing at a variety of LV endocardial (LVENDO) locations. Methods and results: We compiled a registry of acute LVENDO pacing studies from five international centres: Johns Hopkins-USA, Bordeaux-France, Eindhoven-The Netherlands, Oxford-United Kingdom, and Guys and St Thomas' NHS Foundation Trust, London-UK. In all, 104 patients incorporating 687 endocardial and 93 epicardial pacing locations were studied. Mean age was 66 ± 11 years, mean left ventricular ejection fraction 24.6 ± 7.7% and mean QRS duration of 163 ± 30 ms. In all, 50% were ischaemic [ischaemic cardiomyopathy (ICM)]. Scarred segments were associated with worse haemodynamics (dp/dtmax; 890 mmHg/s vs. 982 mmHg/s, P < 0.01). Delivering BiVENDO pacing in areas of electrical latency was associated with greater improvements in AHR (P < 0.01). Stimulating late activating tissue (LVLED >50%) achieved greater increases in AHR than non-late activating tissue (LVLED < 50%) (8.6 ± 9.6% vs. 16.1 ± 16.2%, P = 0.002). However, the LVENDO pacing location with the latest Q-LV, was associated with the optimal AHR in just 62% of cases. Conclusions: Identifying viable LVENDO tissue which displays late electrical activation is crucial to identifying the optimal BiVENDO pacing site. Stimulating late activating tissue (LVLED >50%) yields greater improvements in AHR however, the optimal location is frequently not the site of latest activation.