Reverse remodelling in patients with cardiac resynchronisation therapy-defibrillator: comparison of a conventional 3-lead CRT-D and a 2-lead CRT-DX system with atrial sensing without atrial lead

Background The role of atrial pacing support is unclear in patients with cardiac resynchronisation therapy-defibrillator (CRT-D) and no sinus node dysfunction receiving optimised medical therapy. We conducted a multicentre, randomised, non-inferiority trial (CRT-NEXT) to evaluate whether a two-lead CRT-DX system capable of atrial sensing (but no pacing) via a floating dipole on the right ventricular lead is not inferior to a three-lead CRT-D with conventional atrial lead. Purpose The aim of the present analysis was to compare the reverse remodelling in CRT-DX and CRT-D patients at 12 months. Methods Patients with standard CRT-D indication, optimised medical therapy, no sinus node dysfunction, and resting sinus rate ≥45 beats/min were randomised 1:1 to CRT-DX (VDD mode at 35 beats/min) or CRT-D (DDD, 50 beats/min). At 12 months, echocardiography was used to assess changes in left ventricular ejection fraction (LVEF) and in left ventricular end-diastolic (LVEDV) and end-systolic (LVESV) volumes indexed to body surface area (BSA). The linear mixed-effect models were used with random intercepts at the patient level. A positive response to cardiac resynchronisation therapy was defined as either a ≥5% absolute increase in LVEF or a ≥15% relative decrease in LVEDV. Results 636 patients (age: 68 ± 10 years, 71% males, 41% ischaemic aetiology) were randomised, and 585 patients (92%) reached the 12-month follow-up. LVEF significantly improved (p<0.001) from 29% ± 6% at baseline to 42% ± 9% (CRT-DX) and 41% ± 10% (CRT-D) with no differences between study arms. LVESV/BSA was significantly larger in the CRT-DX group at baseline (74 ± 30 mL/m2) than in the CRT-D group (68 ± 27 mL/m2; p=0.039) and became similar at 12 months (48 ± 26 vs. 47 ± 25 mL/m2, respectively; p=0.43 between groups; p<0.0001 compared to baseline). There was a significant treatment effect favouring CRT-DX (mixed-effect model for LVESV/BSA, p=0.030) (Figure). Likewise, LVEDV/BSA (in mL/m2) was 101 ± 37 (CRT-DX) vs. 95 ± 33 (CRT-D) at baseline (p=0.054), and 77 ± 30 (CRT-DX) vs. 76 ± 30 (CRT-D) at 12 months (p=0.36 between groups; p<0.0001 compared to baseline). There was a non-significant trend in treatment effect favouring CRT-DX (p=0.23). Responder rates were similar in the two groups: 77% (CRT-DX) vs. 76% (CRT-D) (p=0.83). Conclusion At 12 months, CRT-DX and CRT-D systems achieved similar indexed LVEDV and LVESV values despite a significant difference in baseline volumes. Reverse remodelling may be favored by complete suppression of atrial pacing in the selected population.