Effect of PR interval prolongation on long-term outcomes in patients with left bundle branch block vs non-left bundle branch block morphologies undergoing cardiac resynchronization therapy

BACKGROUND Although the influence of QRS duration (QRSd) and/or bundle branch block morphology on outcomes of cardiac resynchronization therapy (CRT) have been well studied, the effect of PR interval remains uncertain. OBJECTIVE The purpose of this study was to evaluate the impact of PR prolongation (PRp) before CRT on long-term outcomes, specifically taking into account bundle branch block morphology and QRSd. METHODS We extracted clinical data on consecutive patients undergoing CRT. Multivariate models were constructed to analyze the effect of PRp (>= 200 ms) on the combined endpoint of death, heart transplant, or left ventricular assist device. Kaplan-Meier curves were constructed stratifying patients based on bundle branch block and QRSd (dichotomized by 150 ms). RESULTS Of the 472 patients who met inclusion criteria, 197 (41.7%) had PR interval >= 200 ms. During follow-up (mean 5.1 +/- 2.6 years) there were 214 endpoints, of which 109 (23.1%) occurred in patients with PRp. In multivariate analysis, PRp was independently associated with worsened outcomes (hazard ratio 1.34, 95% confidence interval 1.01-1.77, P = .04). When stratified by bundle branch block morphology, PRp was significantly associated with worsened outcomes (log-rank P <.001) in patients with LBBB but not in those with non-LBBB (log-rank P = .55). Among patients with LBBB, stratified by QRSd, patients without PRp had improved outcomes compared to those with PRp independent of QRSd (log-rank P <.001). CONCLUSION PRp is an independent predictor of impaired long-term outcome after CRT among patients with LBBB but not in non-LBBB patients. Notably, among LBBB patients, PRp is a more important predictor than QRSd in assessing long-term outcomes.