Factors Associated With Response to Cardiac Resynchronization Therapy in Heart Failure Patients With Non-LBBB ECG Pattern

Morbidity, mortality, and health care costs of the treatment of systolic heart failure (HF) are rapidly increasing. Cardiac resynchronization therapy with a defibrillator (CRT-D) is cost-effectively reducing HF events and death in HF patients with a wide QRS and low ejection fraction. However, response to CRT-D is not unequivocally present in all patients, with less benefit in those without the presence of an ECG abnormality, left bundle branch block (non-LBBB), posing a significant treatment challenge.

Because of the conflicting and limited data on response to CRT-D in this cohort, it is possible that we currently treat a large proportion of HF patients with non-LBBB who have limited or no benefit from the device. Therefore, better selection of patients for this expensive therapy is warranted. One of our recent studies suggested a clinical benefit in patients with non-LBBB and marked echocardiography response, and identified predictors. However, there is a need to prospectively validate these predictors of echocardiography response to CRT-D in non-LBBB in this hard-to-treat patient population, and identify potential novel ECG and echocardiography predictors, utilizing novel statistical methods of machine learning. We propose a prospective, observational, single-arm study in a currently guideline-indicated cohort to validate and identify predictors of echocardiography response to CRT-D, including novel ECG and echocardiography markers, and to assess subsequent clinical outcomes in 270 HF patients with an implanted CRT-D and non-LBBB ECG pattern.

The primary aim of the study is to prospectively validating our previously identified clinical predictors of echocardiography response to CRT-D in HF patients with non-LBBB that could enable better patient selection.

Our secondary aim is to identify the incremental value of novel ECG and echocardiography variables to predict echocardiography response to CRT-D in non-LBBB patients, including ECG variables of sum absolute QRST integral and ventricular electrical activation delay, and echocardiography-derived variables of left ventricular dyssynchrony and contractility. Then we will apply the developed predictive model to prospectively identify non-LBBB patients with CRT-D at a risk of heart failure, ventricular arrhythmias, or death. Tertiary aim is to identify novel ECG and echocardiography predictors of response in non-LBBB using machine learning analysis.

Study population will include 270 HF patients with non-LBBB and an implanted CRT-D with 6 months echocardiography follow-up analyzed by an echocardiography core lab, and assessing clinical outcomes of heart failure, ventricular arrhythmias, or death.

Inclusion criteria

• Age 18 years or older (no upper limit)
• Optimal medical therapy for heart failure as established by current guidelines;
• Class IIa or IIb guideline-based indication for CRT-D in non-LBBB patients, including one of the following:
• New York Heart Association (NYHA) class II HF symptoms, LVEF ≤ 30% and QRS≥ 150 ms (IIb);
• NYHA class III-IVa HF, LVEF ≤ 35%, and QRS duration ≥ 150 ms (IIa);
• NYHA class III-IVa HF, LVEF ≤ 35%, and QRS duration 120- 149 ms (IIb)
• Successful CRT-D generator implant

Exclusion criteria

• Unable to obtain imaging data from echocardiogram within 1 year prior to CRT-D implant
• Unable to obtain the 20-minute ECG prior to CRT-D implant
• Unable to enroll within 14 days following successful CRT-D generator implant
• Unable or unwilling to follow the study protocol
• Less than 24 months life expectancy at enrollment
• Pregnancy or planned pregnancy in the next 24 months