Left bundle branch block and ventricular dysfunction: the importance of QRS duration over strict morphological criteria; data from the RECOrd-LBBB registry

Abstract

Abstract Background Left bundle branch block (LBBB) has been associated with left ventricular (LV) remodeling and dysfunction. LBBB may appear as an ominous sign in the natural course several heart diseases, or may play a causative role in the development of a “dyssynchronopathy” (LBBB-induced cardiomyopathy). LBBB represents an important therapeutic target in heart failure (HF) patients, since cardiac resynchronization therapy (CRT) aims to restore mechanical synchrony. Over the last 10 years different electrocardiographic (ECG) criteria (by the American Heart Association – AHA, and by Strauss) have been proposed to better identify and select patients suitable for CRT. The real usefulness of these criteria has been recently questioned. Purpose Aims of this single center registry were: 1) to describe the clinical and instrumental (ECG and echocardiographic) features in patients with LBBB hospitalized due to both cardiac and non-cardiac causes in a tertiary university hospital; 2) to identify the prevalence of LV remodeling/dysfunction, and its potential determinants. Methods One hundred eighty-six patients underwent clinical evaluation, 12-lead ECG, and transthoracic echocardiogram. Exclusion criteria were: 1) prior myocardial infarction with extensive anterior necrosis; 2) prior cardiac surgery or transcatheter aortic valve implantation; 3) complex congenital heart diseases; 4) permanent pacing with right ventricular stimulation. Results One hundred and three (55%) were males. Median age was 74.5 (IQR 66–82) years. Arterial hypertension was the most prevalent (65%) cardiovascular risk factor. Twenty-one patients had atrial fibrillation. Median QRS duration was 140 (IQR 140–160) msec, 70% of patients had a LBBB according to the AHA criteria, while 85% was diagnosed as having a “true” LBBB according to Strauss definition. A discordant LBBB was found in 56% of patients. Median LV ejection fraction (EF) was 42%, and LV systolic dysfunction was found in 66% of patients. There was a weak, although statistically significant, inverse correlation between QRS duration (msec) and LVEF (rho −0.23; p=0.0026). There was a significant difference in terms of QRS duration (160 vs 140 msec; p=0.008) between patients with and without LV dysfunction, while LBBB morphological criteria did not differentiate the two groups. QRS duration resulted to be the only independent determinant of LV systolic dysfunction (OR 1.022 [95% CI]=1.022 [1.002–1.042], p=0.029) on multivariable logistic regression analysis. Conclusions In a non-selected cohort of patients with LBBB, the use of strict morphological criteria does not predict the presence of LV dysfunction. QRS duration seems to be the only independent predictor of LV dysfunction. These results emphasize the complex relationship between LBBB and cardiac remodeling and dysfunction. Funding Acknowledgement Type of funding sources: None.