Activation time at left ventricular pacing site (QLV) in relation to actual site of latest activation; effect on response to cardiac resynchronization therapy

Abstract

Abstract Background Among techniques to improve CRT efficacy, longer QLV (activation time at the left ventricular (LV) pacing site (LVPS)) has been proposed to correlate to response to cardiac resynchronization therapy (CRT), supporting that pacing close to the latest LV electrically activated site (LEAS) improves CRT efficacy. A QLV cut-off of > 95 ms has been proposed, however, predictive value was modest. This may be because LEAS was not identified, and empirically delivered leads may have missed it (described by a longer interval QLEAS (activation time at LEAS) to QLV. Purpose We compared QLEAS to QLV, and assessed effect of the interval QLEAS-QLV, on response to CRT. Methods We studied patients who had received CRT for Class I or Class IIa indications, regardless of CRT response, at 5 European sites. LV lead placement, device programming and follow up (FU) followed physician preference and site protocol. All patients had clinical FU 6-12 months post-implant, including echocardiographic study. CRT non-response was defined as LV end-systolic volume (LVESV) reduction of <15%. Patients underwent noninvasive 3D electrical activation mapping and torso CT 6-24 months post-implant. QLV and QLEAS were assessed from the noninvasive activation map in a blinded core lab. Logistic regression modelling was conducted for QLV and the interval QLEAS-QLV against CRT response, and the optimal cut-off points were calculated. Results Of 111 patients (age 64±11 years (mean±SD), 74% male, 98% LBBB, QRS duration 172±21 ms, baseline LVEF 28±6%, LVESV 183±87 mL, implant duration 12±5 months, 31%/69% NYHA Class II/III, 38% ischemic heart disease), 67% responded at 10±3 months post-implant. QLV overall was 97±23 ms (range 42-148 ms), 103±21 ms for responders and 83±22 for non- responders (p<0.001), QLEAS was 109±19 ms (range 66-150 ms), 109±20 ms for responders and 110±19 ms for non- responders (p=0.74). QLEAS-QLV overall was 13±18 ms (range 0-120 ms), 6±10 ms for responders and 27±22 ms for non-responders (p<0.001) (Figure 1). Optimal cut-off point for QLV was 85 ms with sensitivity 0.89, specificity 0.57 and an AUC of 0.742 (95% CI: 0.637- 0.846). Optimal cut-off point for QLEAS-QLV was 6 ms (sensitivity 0.77, specificity 0.95, AUC 0.901 (95% CI: 0.844-0.957) (Figure 1). The AUC of QLV cut-off and QLEAS-QLV cut-off differed significantly (p<0.001). 18/111 patients (16%) were non-responders despite a QLV≥85 ms, whereas 2 patients (2%) were non-responders despite a QLEAS-QLV≤6 ms. Conclusion QLEAS has greater discriminatory power than a QLV cut-off of 85 ms to separate CRT responders from non-responders in patients with LBBB and long QRS duration. QLV guided LV lead placement without prior knowledge of the latest LV activation in individual patients will miss the ideal LV pacing site in 16% of cases. Using noninvasive 3D global mapping to identify LEAS pre-CRT implant and directing LV lead placement to within 6 ms of this may improve CRT efficacy.Figure 1