Relationship of myocardial substrate characteristics as assessed by myocardial perfusion imaging and cardiac reverse remodeling levels after cardiac resynchronization therapy

Background Cardiac resynchronization therapy (CRT) can provide cardiac reverse remodeling (RR), which may include mechanical reverse remodeling (MRR) and/or electrical reverse remodeling (ERR). However, uncoupling of MRR and ERR is not uncommon, and the underlying mechanisms are not clear. This study aimed to evaluate the relationship of myocardial substrate characteristics as assessed by myocardial perfusion imaging (MPI) and cardiac RR post-CRT. Materials and methods Forty-one patients (26 men, mean age 66 ± 10 years) with heart failure received CRT for at least 12 months were assigned to three groups according to their levels of RR: I, MRR + ERR (ESV reduced ≥15 % and intrinsic QRS duration reduced ≥10 ms); II, MRR only (ESV reduced ≥15 %); and III, non-responder (the others). All the patients also underwent MPI under transient CRT-off to evaluate the intrinsic myocardial substrates, including myocardial scar, LV volumes and function, systolic dyssynchrony, and activation sequences. In addition, ventricular tachycardia (VT) and ventricular fibrillation (VF) detected by the CRT devices during follow-up periods were also recorded. Results Quantitative analysis of MPI showed that there were significant differences for scar burden [15.9 ± 9.5, 26.8 ± 16.1, and 45.6 ± 15.1 % for group I ( n = 15), II ( n = 16), and III ( n = 10), respectively, p < 0.001], EDV (136.6 ± 64.9, 221.6 ± 123.9, and 351.8 ± 216.3 ml, p = 0.002), ESV (82.6 ± 59.8, 172.3 ± 117.2, and 293.3 ± 209.6 ml, p = 0.001), LVEF (44.9 ± 15.0, 25.6 ± 10.9, and 21.5 ± 11.7 %, p < 0.001), systolic phase SD (23.4° ± 10.3°, 36.0° ± 16.2°, and 57.0° ± 22.2°, p < 0.001), and bandwidth (72.5° ± 31.1°, 113.4° ± 56.4°, and 199.1° ± 90.1°, p < 0.001). Myocardial scar interfered with the normal propagation of mechanical activation, resulting in heterogeneous activation sequences. Compared with group II (MRR only), group I (ERR + MRR) had significantly less initial activation segments (1.9 ± 1.0 vs. 2.6 ± 0.7, p < 0.05) and shorter maximal contraction delay (46.9° ± 12.9° vs. 58.8° ± 18.5°, p < 0.05). During the periods of follow-up, 21 patients developed VT/VF, including only 1 patient (1 VT) in group I (6.7 %), 8 patients (7 VT and 1 VF) in group II (50 %), and 9 patients (7 VT and 5 VF) in group III (90 %). Conclusion The characteristics of myocardial substrates as assessed by MPI differed significantly between different levels of cardiac RR post-CRT. Myocardial scar played an important role in the development of ERR. Different cardiac RR levels contributed to different incidences of ventricular arrhythmia, and the combination of ERR and MRR provided highest anti-arrhythmic effects.