Effect of QRS area reduction and myocardial scar on the hemodynamic response to cardiac resynchronization therapy

Background

Vectorcardiographic QRS area (QRS_area) predicts clinical outcomes after cardiac resynchronization therapy (CRT). Myocardial scar adversely affects clinical outcomes after CRT.

Objective

The purpose of this study in patients with an ideally deployed quadripolar left ventricular (LV) lead (QUAD) was to determine whether reducing QRS_area leads to an acute hemodynamic response (AHR) and whether scar affects this interaction.

Methods

Patients (n = 26; age 69.2 ± 9.12 years [mean ± SD]) underwent assessment of the maximum rate of change of LV pressure (ΔLV dP/dt_max) during CRT using various left ventricular pacing locations (LVPLs). Cardiac magnetic resonance (CMR) scan was used to localize LV myocardial scar.

Results

Interindividually, ΔQRS_area (area under the receiver operating characteristic curve [AUC] 0.81; P <.001) and change in QRS duration (ΔQRSd) (AUC 0.76; P <.001) predicted ΔLV dP/dt_max after CRT. Scar burden correlated with ΔQRS_area (r = 0.35; P = .003), ΔQRS_area (r = 0.35; P = .003), and ΔQRSd (r = 0.46; P <.001). A reduction in QRS_area was observed with LVPLs remote from scar (–3.28 ± 38.1 μVs) or in LVPLs in patients with no scar at all (–43.8 ± 36.8 μVs), whereas LVPLs over scar increased QRS_area (22.2 ± 58.4 μVs) (P <.001 for all comparisons). LVPLs within 1 scarred LV segment were associated with lower ΔLV dP/dt_max (–2.21% ± 11.5%) than LVPLs remote from scar (5.23% ± 10.3%; P <.001) or LVPLs in patients with no scar at all (10.2% ± 7.75%) (both P <.001).

Conclusion

Reducing QRS_area improves the AHR to CRT. Myocardial scar adversely affects ΔQRS_area and the AHR. These findings may support the use of ΔQRS_area and CMR in optimizing CRT using QUAD.

Introduction

Cardiac resynchronization therapy (CRT) is an established therapy for patients with impaired left ventricular (LV) systolic function and a wide QRS complex. In addition to improving symptoms and quality of life, CRT reduces heart failure hospitalizations and improves survival.¹ However, even in guideline-indicated patients, the response to CRT is variable. The nonresponder rate ranges between 9% and 68%, depending on the criteria used to define response.² Prominent among the factors implicated in nonresponse is suboptimal LV lead deployment.

Quadripolar LV leads (QUAD) have been a game-changer in the field of CRT. The availability of a wide range of left ventricular pacing locations (LVPLs) not only reduces diaphragmatic stimulation but also permits LV pacing from widely spaced areas of the myocardium. Several studies have shown that optimizing LVPLs on a QUAD in individual patients improves the acute hemodynamic response (AHR) to CRT.^3,4 Myocardial scar in the vicinity of the LVPL has been shown to be detrimental in studies using bipolar leads,^5,6 but no studies have explored whether the same applies to QUAD.

It has been shown recently that vectorcardiography (VCG)-derived QRS area (QRS_area) is a measure of LV electrical dyssynchrony, the natural substrate of CRT.⁷ Moreover, QRS_area has been shown to correlate with AHR^8,9 and LV reverse remodeling¹⁰ after CRT. A reduction in QRS_area has been shown to improve AHR to CRT in patients with suboptimal LV lead deployment.¹¹ Importantly, preimplant QRS_area has been shown to be superior to QRS duration (QRSd) in predicting total mortality after CRT.^12,13 We recently showed that a postimplant reduction in both QRS_area and QRSd was associated with the best outcomes after CRT.¹⁴ In the present study, we explored whether in patients with an optimally deployed LV lead assessed by fluoroscopy, a change in QRS_area (ΔQRS_area) after CRT relates to AHR; AHR in individual patients can be improved by reducing QRS_area; and AHR relates to the position of the LVPL in relation to myocardial scar, assessed using cardiac magnetic resonance (CMR).