Septal flash correction with His-Purkinje pacing predicts echocardiographic response in resynchronization therapy

Abstract

Background: His-Purkinje conduction system pacing (HPCSP) has been proposed as an alternative to Cardiac Resynchronization Therapy (CRT); however, predictors of echocardiographic response have not been described in this population. Septal flash (SF), a fast contraction and relaxation of the septum, is a marker of intraventricular dyssynchrony.

Methods: The study aimed to analyze whether HPCSP corrects SF in patients with CRT indication, and if correction of SF predicts echocardiographic response. This retrospective analysis of prospectively collected data included 30 patients. Left ventricular ejection fraction (LVEF) was measured with echocardiography at baseline and at 6-month follow-up. Echocardiographic response was defined as increase in five points in LVEF.

Results: HPCSP shortened QRS duration by 48 ± 21 ms and SF was significantly decreased (baseline 3.6 ± 2.2 mm vs. HPCSP 1.5 ± 1.5 mm p < .0001). At 6-month follow-up, mean LVEF improvement was 8.6% ± 8.7% and 64% of patients were responders. There was a significant correlation between SF correction and increased LVEF (r = .61, p = .004). A correction of ≥1.5 mm (baseline SF - paced SF) had a sensitivity of 81% and 80% specificity to predict echocardiographic response (area under the curve 0.856, p = .019).

Conclusion: HPCSP improves intraventricular dyssynchrony and results in 64% echocardiographic responders at 6-month follow-up. Dyssynchrony improvement with SF correction may predict echocardiographic response at 6-month follow-up.

Keywords: His-Purkinje conduction system pacing; intraventricular dyssynchrony; left ventricular ejection fraction; physiological pacing; septal flash.

FIGURE 1

Central Illustration. Correction of septal flash (SF) with His Purkinje conduction system pacing (HPCSP). (A) Left bundle branch block with QRS width of 196 ms and SF of 5 mm; electrocardiographic imaging showed a ventricular activation pattern with late activation of the lateral left ventricle (in blue) and a left ventricular activation time (LVAT) of 82 ms. (B) HPCSP with left bundle branch pacing with AV optimization (AV 90 ms) obtained QRS of 124 ms, abolition of SF, and fast ventricular activation (red) with LVAT of 41 ms. (C) The adjusted probability prediction for LVEF response showed that the probability of response was 86% if SF correction was ≥1.5 mm

FIGURE 2

Left bundle branch pacing (LBBP) optimization with fusion. Early activation shown in red, late activation in blue. (A) Baseline activation with left bundle branch block (QRS 196 ms); electrocardiographic imaging showed ventricular activation pattern with late activation of the lateral left ventricle (in blue). (B) LBBP with short AV (60 ms), achieving a QRS of 146 ms. (C) Prolongation of the AV delay allows coordinating the stimulation of the left branch with the intrinsic activation of the right branch, normalizing the activation time and pattern (both ventricles in red); AV 90 ms allows the shortest QRS (124 ms). (D) With AV 120 ms, right ventricle is preactivated (red), obtaining a QRS of 132 ms

FIGURE 3

Septal flash correction with His Purkinje conduction system pacing. Activation in patients with left bundle branch block produces septal flash (SF), a marker of intraventricular dyssynchrony. Black arrows point to SF in two patients (A and B), a fast contraction and relaxation of the septum occurring during the isovolumetric contraction period. SF was corrected (yellow arrow, A) or abolished (yellow arrow, B) with His bundle pacing [Colour figure can be viewed at wileyonlinelibrary.com]

FIGURE 4

Baseline and physiological pacing septal flash (SF). Representation of SF values at baseline and with physiological pacing, in each patient (n = 30) (A) and Box plot diagram (B) [Colour figure can be viewed at wileyonlinelibrary.com]

FIGURE 5

(A) Increase of left ventricular ejection fraction (LVEF) as a function of delta Septal flash. Correction of SF correlated with LVEF increase at 6 months (r = .61, p = .004). (B) Receiver operating characteristic (ROC) curve illustrates the classifying ability of SF correction to predict echocardiographic response. A value of SF correction ≥1.5 mm had 81% sensitivity and 80% specificity to predict echocardiographic response at 6‐month follow‐up (area under the curve 0.856, p = .019)