High predictive value of paced QRS frequency in verification of left bundle branch pacing

Left bundle branch area pacing (LBBAP) is a newer technique to deliver more synchronous left ventricular activation. Several criteria have been proposed, but not fully validated, to confirm LBBAP during implantation of the pacing lead. Spectral analysis has been used to characterize the frequency components of the clinical QRS utilizing the Fourier transform algorithm. We hypothesized that higher frequency content of the paced QRS complex may show predictive value of successful LBBAP.

We evaluated 84 patients with ejection fraction > 50%, who underwent LBB lead placement (n = 42) using ≥ 1 current criteria and right ventricular midseptal (RVsp) lead placement (n = 42) from 2000 to 2022. Time frequency analysis (Matlab) was used to determine the frequency content of the paced QRS complex. The centroid frequency (CF), which is the weighted average QRS frequency, was calculated.

Patients in RVsp group had a longer paced QRS duration (155.6 ± 28.0 vs 127.1 ± 17.2, p < 0.002) compared to the LBBAP group. Of all standard ECG leads, the paced QRS in V2 gave the greatest difference of the CF of the LBBAP group at 8.8 ± 1.6 Hz versus 5.7 ± 0.7 Hz of the RVsp group. This difference was significant by both univariate (p < 0.003) and multivariate (p < 0.010) analysis. Predictive value of the CF for successful LBB pacing in lead V2 was highest with an AUC of 0.98. The sensitivity and specificity were 88.1% and 97.6%, respectively.

Spectral analysis predicts successful LBBAP with higher frequency content when compared to RVsp pacing. Given the limitations to the current criteria to confirm LBBAP, intraprocedural use of frequency content analysis of the paced QRS complex in patients may prove useful at verifying LBB capture if verified by prospective clinical trials.