Analysis of long-term mortality outcomes in cardiac resynchronization therapy

Even in well-selected patients with drug-refractory heart failure, 20-40% remain non-responders to cardiac resynchronization therapy (CRT). Although the criteria for non-response are heterogeneous across studies and may refer to clinical, echocardiographic or functional parameters, studies show an association with increased mortality. There is substantial evidence regarding short- and mid-term response after CRT, however, data regarding mortality outcomes over 5 years of follow-up remain scarce.

The primary objective of this study is to assess long-term survival after CRT. The secondary aim is to evaluate and identify factors associated with all-cause mortality over the follow-up period.

This retrospective cohort study included 213 patients who underwent de novo CRT implantation according to available guidelines at the time of the procedure, between 2010 and 2020. Exclusion criteria consisted of other indications for pacing, device upgrades, permanent atrial fibrillation and insufficient data.

After a mean follow-up of 71.6 months, 93 (43.7%) patients died, with an observed cumulative mortality of 6.6% at 1 year, 15% at 2 years, 23.5% at 3 years, 38.5% at 5 years and 88.3% at 10 years. CRT-P population was dominant, with 157 (74%) patients, and device type was not correlated with differences in survival. Mortality in the first year was driven by cardiovascular causes, including arrhythmias, such as ventricular tachycardia or even electrical storm, which represented 36% of deaths, and heart failure in 21% of cases. Arrhythmic deaths represented 35.7% of first-year mortality, significantly higher than in patients who died between 1 and 5 years (6.9%) or after 5 years (14.3%) post-CRT (Bonferroni-adjusted p=0.01). No other differences in cause of death were found between the three groups. Two-thirds of patients who presented at the index admission with decompensated heart failure died within five years, as well as almost half of the patients with ischaemic heart failure. Approximately one half of all observed deaths were non-cardiac. All-cause mortality was correlated with older implantation age, anticoagulation, larger baseline ventricular volumes, lower baseline ejection fraction, NYHA class III-IV, moderate and severe tricuspid or mitral regurgitation or chronic kidney disease.

Early mortality is driven by arrhythmic burden, ischaemia and heart failure severity at the index presentation, whereas factors such as heart failure progression and comorbidities influence mid- and long-term outcomes.