Background: Patients with persistent atrial fibrillation (AF) may experience reduced left ventricular ejection fraction (LVEF) despite rate-control. The non-normal distribution of R–R intervals may be poorly represented by the mean heart rate and a non-parametric measure of the proportion of short R–R intervals may more sensitively characterize the AF-mediated component of left ventricular systolic dysfunction.
Purpose: To evaluate the restitution threshold index (RTI); a novel parameter that reflects the proportion of short R–R intervals in patients with AF with reduced and preserved LVEF.
Methods: Patients undergoing first-time catheter ablation for persistent AF were prospectively enrolled. Patients were dichotomised as reduced versus preserved LVEF using an LVEF of 50% as cut-off. Holter monitoring was performed prior to the ablation with a supervised and standardised initial recording period to minimize confounders. The Restitution Threshold (RT) refers to the assumed minimum cycle length duration required for complete mechanical restitution and filling of the left ventricle. This threshold is unknown and was tested in both cohorts across a range of 300 ms and 1200 ms at 10 ms increments. A high proportion of R–R intervals below the actual RT could lead to acute LVSD and impaired contractility. This proportion of intervals below the RT was defined as the RTI. Multi-variate logistic regression modelling was used to assess the relationship between the RT with maximum separation of RTI between the two cohorts and pre-defined co-variates (mean heart rate, indexed left atrial volume, age, sex, beta-blocker and amiodarone use)
Results: 46 patients were included, with 22 in the preserved LVEF arm and 24 in the reduced LVEF arm. There was significant difference in RTI between the two arms at all RTs between 450–900 ms. (Figure 1) Maximum separation was observed at an RT of 650 ms, with patients with reduced LVEF exhibiting a significantly higher RTI (60.5% ± 24.7% versus 36.8% ± 24.3%, p=0.002). RTI was an independent predictor of patients with reduced LVEF on multi-variate analysis (B=0.038, p<0.01), whereas mean heart rate was not.
Conclusions: This exploratory study demonstrates feasibility of RTI as a novel parameter characterising the proportion of short R–R intervals may better identify patients with LVSD in persistent AF, with an RT of 650 ms performing best in our cohort. RTI may have potential clinical application in selecting patients with tachycardia-induced cardiomyopathy, but further validation is required. ❑
Figure 1: Mean restitution threshold index distribution
The distribution of R–R intervals in patients with reduced left ventricular (LV) function (blue) and preserved LV function (orange) during persistent AF plotted as mean (line) and two standard errors of the mean (shaded area). This is shown for a range of values for the RT. The dotted line represents the point of maximum separation at 650 ms.
