Does Paced QRS Duration Predict the Risk of Heart Failure Events During Permanent Right Ventricular Pacing?

This editorial refers to ‘Paced QRS duration as a predictor for clinical heart failure events during right ventricular apical pacing in patients with idiopathic complete atrioventricular block: results from an observational cohort study (PREDICT-HF)’, by S. Chen et al., published in this issue on pages 352–359. In this issue, Chen et al. report observational data1 suggesting that paced QRS duration (pQRSd) may predict the risk of new or worsening heart failure in patients implanted with a conventional pacemaker for high-degree atrioventricular (AV) block, resulting in permanent right ventricular capture. Animal and human studies showed that right ventricular apical (RVA) pacing may cause detrimental ventricular remodelling. This negative effect has been attributed to the abnormal electrical activation and to dyssynchronous mechanical contraction/relaxation within the left ventricle as observed for ectopic stimulation.2,3 Abnormal prolongation of the pQRSd reflects electrical dyssynchrony. Pacing-induced mechanical dyssynchrony is associated with asymmetric hypertrophy, left atrium enlargement, mitral regurgitation, and decreased LVEF.4,5 The incidence of new or worsening heart failure in patients with permanent RVA pacing has been assessed in observational and randomized studies, ranging from 10% to 26%.6,7 Zhang et al. studied the clinical outcomes after long-term RVA pacing for acquired AV block of 304 patients without history of heart failure. After a median follow-up of 7.8 years, the authors observed that 26% of patients developed symptoms or signs of heart failure.7 In the MOST (Mode Selection Trial) study, a randomized controlled trial on pacing mode selection in patients with sinus node dysfunction, a strong association was observed between the percentage of ventricular pacing and the risk of heart failure hospitalization.8 The most important predictive factor was the percentage of ventricular pacing regarding the prevention of heart failure events.5 Nevertheless, the risk of new or worsening heart failure in patients submitted to RVA pacing is primarily related to the cardiac substrate. Sweeney et al. reported that patients with decreased LVEF, history of symptomatic heart failure, history of myocardial infarction, or native prolonged QRS duration had an increased risk of heart failure hospitalization.5 In summary, patients with structural heart disease and LV dysfunction are at higher risk of developing heart failure during RVA pacing, and efforts to minimize pacing might be considered whenever possible, particularly in this group of patients. However, permanent ventricular capture cannot be avoided in a large proportion of pacemaker patients, namely those implanted for high-degree AV block. In this population, risk stratification based on pre- and perioperative patient characteristics may serve to identify patients at risk of developing heart failure after RVA pacing and eventually to consider alternative pacing strategies. Though the clinical evidence is weak, the European Society of Cardiology (ESC) guidelines for the diagnosis and treatment of acute and chronic heart failure 2012 have issued a class IIa (Level of evidence C) consensus recommendation to consider implantation of a CRT device and thus to avoid RVA pacing in patients with conventional pacemaker indication, NYHA class III–IV heart failure, and LVEF <35%, irrespective of the intrinsic QRS duration.9 A class IIb (Level of evidence C) recommendation was also issued for NYHA class II patients with a similar presentation.9 To date, no major randomized controlled trials on CRT have included patients with a conventional pacemaker indication, except for RAFT,10 but the number of patients enrolled in this subgroup was too small to detect statistical significance. Regarding perioperative characteristics, Chen et al. report the results of the PREDICT-HF study aimed at assessing the predictive value of pQRSd.1 This observational, single-centre study investigates the predictive value of pQRSd on the risk of heart failure events among 194 patients (mean age 66.7 ± 12.6 years; 96 men) with high-degree AV block over a 3-year follow-up period. The patients had normal native QRS morphology, narrow QRS duration (<120 ms), and no history of heart failure. Patients were stratified into three groups according to the pQRSd measured after implantation: group 1, pQRSd ≤ 160 ms (n = 53 patients); group 2, 160 < pQRSd ≤ 190 ms (n = 97 patients); and group 3, pQRSd ≥ 190 ms (n = 44 patients). The incidence of heart failure events was 9.4, 27.8, and 56.8% in groups 1, 2, and 3, respectively (P < 0.001). A receiver operating characteristic (ROC) analysis determined a cut-off point of 165 ms for long-term risk of heart failure with a sensitivity of 0.789 and a specificity of 0.679. Furthermore, a decrease in LVEF over the time was positively correlated with pQRSd (relative risk 0.423, P < 0.05). The present study was not the first to investigate the correlation between pQRSd and the long-term changes in heart failure status as assessed by symptoms, LV size [LV diastolic and LV diastolic dimensions), and function (LVEF), and/or biomarkers.11 In a post-hoc analysis of MOST, Shukla et al. reported that pQRSd was a significant, independent predictor of heart failure hospitalization in patients with sinus node dysfunction implanted with a VVIR or DDDR pacemaker [hazard ratio (HR) 1.15; 95% confidence interval (CI) 1.07–1.23 for each 10 ms increase in pQRSd].12 The wider the pQRSd, the greater was the risk of heart failure events. pQRSd might be an earlier marker of LV function even if LVEF is preserved. However, the value of pQRSd in predicting heart failure events in patients implanted for sinus node dysfunction is uncertain as AV delay may influence the pQRSd with a certain level of fusion beats. Miyoshi et al. studied the value of pQRSd in predicting congestive heart failure in 92 patients with high-degree AV block.13 pQRSd was positively correlated with the changes in LV dimensions and function, and a pQRSd >190 ms was associated with a significant increase in the risk of developing heart failure. In the PREDICT-HF study, Chen et al. report a progressive increase in pQRSd over time. At 3 years after pacemaker implantation, change in pQRSd was greater in patients with heart failure events as compared with patients without hear failure events, suggesting that the change in pQRSd during follow-up might also predict the risk of heart failure events. These findings are consistent with those previously reported by Miyoshi et al. These authors observed that pQRSd significantly increased during a follow-up period of 53 ± 16 months in the only patients developing heart failure.13 Pap et al. observed that native QRS width was a predictor of pQRSd after RVA pacing in 50 patients with AV block and normal LVEF.14 In the present study, there were no correlation between paced and native QRS, and native QRS duration was not predictive of the risk of heart failure events. Nevertherless, it is well established that native QRS duration is correlated with LV dimensions and LVEF, and is an important marker of poor outcome in chronic heart failure.15 In the study of Chen et al., only patients without prolonged native QRS were included, suggesting that pQRSd could be an earlier marker of LV dysfunction than the native QRS. In summary, the study of Chen et al.1 provides some additional evidence that post-implant pQRSd and its change over time may contribute to the identification of RVA paced patients at risk of developing LV dysfunction and heart failure. Additional studies are needed to confirm these preliminary results and to determine the pQRSd value (absolute post-implant and/or percentage change over time) best predicting the risk of heart failure development. As pQRSd is a simple and relatively reproducible measurement in patients with RVA pacing, it could be recommended to be performed at the time of implantation and during follow-up. Patients with a long pQRSd at implant and/or increasing pQRSd during follow-up need special attention. Conflict of interest: none declared.