Methods

Study Population

To be enrolled, patients must have been at least 18 years of age, able to provide informed consent, and indicated for de novo CRT-D device implantation. Patients were excluded who did not exhibit left bundle branch block (LBBB), had an epicardial ventricular lead system implanted, exhibited an intrinsic atrial rate below 40 bpm, exhibited atrial fibrillation (AF), had a life expectancy less than 1 year, were pregnant, or dependent on IV inotropic agents. Patients were classified as exhibiting AF if it was either (i) persistent, (ii) permanent and not treated with AV node ablation within 2 weeks of CRT implant, or (iii) documented as paroxysmal or persistent within 30 days of enrollment.

Study Design

This prospective, chronic, randomized, international study was conducted at 13 centers in the Middle East: National Guard Hospital (Riyadh, KSA), King Fahad Armed Forces Hospital (Jeddah, KSA), Prince Sultan Cardiac Center (Riyadh, KSA), Royal Hospital (Muscat, Oman), Ibn Al-Bitar Cardiac Center (Baghdad, Iraq), Nasiriya Heart Center (Nasiriya, Iraq), King Khalid University Hospital (Riyadh, KSA), King Faisal Hospital and Research Center (Riyadh, KSA), Bahrain Defense Force Hospital (Manama, Bahrain), Aswan Heart Centre (Aswan, Egypt), King Abdulaziz University Hospital (Jeddah, KSA), Cleveland Clinic Abu Dhabi (Abu Dhabi, UAE), and Sabah Al-Ahmad Cardiac Center (Kuwait City, Kuwait). All patients provided informed consent, and all study protocols were approved by the ethics committee of each institution.
Enrolled patients were implanted with a de novo CRT-D device (Abbott Unify Quadra MP™ or Quadra Assura MP™) and quadripolar LV lead (Abbott Quartet™) according to standard practice. Prior to hospital discharge, patients were randomized into one of two treatment groups, as follows.
  1. BiV Group : devices programmed to conventional BiV pacing, with the single LV pacing vector selected according to the standard practice of the implanting physician.
  2. MPP Group : devices programmed with MPP enabled, with the two LV pacing vectors (LV1, LV2) selected using the automatic VectSelectTM feature. Specifically, the VectSelect feature was used to assign LV1 and LV2 cathodes to the LV electrodes with the earliest and latest RV-LV conduction times (RV-sensed to LV-sensed), respectively, with corresponding anodes assigned to yield capture thresholds less than 3.5 V with no phrenic nerve stimulation observed at an output of 1.5x capture threshold.
For both groups, programming of the atrioventricular delay was left to the discretion of the implanting physician. For BiV, the interventricular delay was set to the default value (LVRV, 10 ms). For MPP, both the intraventricular and interventricular delays were set to the default values (LV1LV2, 5 ms; LV2RV, 5 ms).
Echocardiography and 12-lead surface electrocardiography were performed at implant and repeated 6 months post-implant. Echocardiographic metrics were analyzed by blinded core labs. ECG QRS duration (QRSd) was defined as the time from the earliest start-time (departure from isoelectric) to the latest end-time (return to isoelectric) across all ECG leads, following standard recommendations and ignoring any pre-QRS deflections attributed to pacing artifacts.16

Study Endpoints

CRT response was characterized by changes in LV end-systolic volume (ΔESV) and LV ejection fraction (ΔEF) at 6 months post-implant. Patients were classified as ESV responders if they demonstrated a relative reduction in ESV of at least 15% vs. baseline. The primary endpoint of the study was a comparison of the proportion of ESV responders (i.e., ESV responder rate) between the BiV and MPP groups. In addition, a combined ESV+EF response classification was also used, in which patients were classified as ESV+EF responders if they demonstrated both a relative reduction in ESV of at least 10% and an absolute elevation in EF of at least 5% vs. baseline.17 Secondary endpoints included the following comparisons between BiV and MPP groups: ESV reduction, EF elevation, QRS duration (QRSd) reduction, ESV+EF responder rate, and improvement in NYHA functional class.

Statistical Analysis

Statistical analyses were performed using Matlab (Statistics Toolbox, The Mathworks). Categorical variables were reported by patient count and patient percentage. Differences in categorical variables between groups were tested using Fisher’s exact test. Continuous variables were reported as median and interquartile range (IQR), as none demonstrated standard normal distributions according to one-sample Kolmogorov-Smirnov tests. Differences in continuous variables between groups were tested using the Mann–Whitney U test. The impact of baseline characteristics on ESV responder rates was quantified using binomial regression. For all statistical tests, P<0.05 was considered statistically significant.