Can 12 lead ECG predit the level of conduction system disease?

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Introduction Permanent His bundle pacing (p-HBP) could be an alternative for cardiac resynchronization therapy (CRT), but an important limitation is that p-HBP cannot always correct the left bundle branch block (LBBB). The aim is to assess electrocardiographic (ECG) patterns of LBBB that can predict ECG response (QRS narrowing) to p-HBP. Methods Prospective descriptive study of patients with LBBB and CRT indication proposed for CRT by p-HBP. We analyzed the correlation between the different ECG patterns and the correction of LBBB. Results We included 70 patients. Once located His bundle electrogram, pacing at this level narrowed the QRS in 81.4% (n=57). We distinguished two ECG patterns in lead V1: QS or rS. Thirty-four patients (51.4%) presented QS pattern in lead V1 (Fig 1A), and in 94.1% of these correction of the LBBB was achieved (Fig 1B). The rS pattern was present in 36 patients and in 69.4% of these the correction of the LBBB was achieved. Differences between both groups were statistically significant (p=.008). The presence of QS pattern in V1 had a sensitivity of 56%, a specificity of 84.6% and a positive predictive value of 94.1% to predict LBBB correction. The AUC by a receiver operating characteristic analysis was .70±. In patients with rS pattern in lead V1, we analyzed the ratio between the descending and the ascending S wave component duration; the difference between patients in whom LBBB was corrected (ratio 0.742±0.124) and patients in whom it was not corrected (0.476±0.124) were statistically significant (p=.001). The AUC by a receiver operating characteristic analysis was .968 (very good predictor of LBBB correction). In this case, a ratio between the descending and the ascending S wave component duration≥0.64 (Fig 2C-D) showed a sensitivity of 92%, a specificity and a PPV of 100%, and a negative predictive value (NPV) of 83.3%. Related to the ratio between the ascending S wave component duration and the total QRS duration, it was 0.513±0.043 in patients with LBBB correction vs 0.60 ± 0.046 in those in whom it did not correct (P=.001). The AUC by a receiver operating characteristics analysis was .896 (good test to predict LBBB correction in patients with rS pattern in lead V1). We analyzed the ratio between the descending and the ascending S wave component duration in the global of patients, including those with QS pattern. Patients in whom it corrected the LBBB presented a ratio of 0.754±0.13 vs 0.459±0.081 in whom it did not correct (P<.001) (AUC .977). The ratio between the ascending S wave component duration and the total QRS duration in patients with LBBB correction was 0.542±0.003 vs 0.623±0.063 in whom it did not correct (P<.001) (AUC .84). Conclusions In patients with LBBB and CRT indication, the QS pattern in lead V1 predicts the correction of the QRS with HBP. In the case of rS pattern in lead V1, the ratio descending/ascending S wave component duration has a strong correlation with the LBBB correction.