Relationship between QRS duration and resynchronization window for CRT optimization: Implications for CRT in narrow...

Cardiac resynchronization therapy (CRT) response is proportional to QRS duration (QRS_d). We hypothesize that this is, in part, due to slower conduction velocity and hence wider range of programmed device settings that produce adequate electrical wavefront fusion and resynchronization in wider QRS_d patients.

CRT patients (n = 122) with left ventricular (LV) conduction delay, sinus rhythm and intact atrioventricular node conduction were studied. Patients were categorized by QRS_d: narrow (<120 ms; n = 20); moderate (120-150 ms, n = 37); and prolonged (≥150 ms; n = 65). Electrocardiographic data was acquired during native rhythm and LV-only pacing at varying atrioventricular delays (AVDs). Electrical synchrony was quantified as cardiac resynchronization index (CRI) using multi‑lead electrocardiographic systems and a proprietary algorithm that quantified wavefront fusion. A Gaussian distribution equation was fitted to CRI response.

Peak CRI was high (87.6 ± 6.3%) and similar (p = 0.716) across QRS_d groups. The standard deviation of the Gaussian distribution significantly correlated with QRS_d (R = 0.614, p < 0.001), and progressively and significantly (p < 0.001) increased as QRS_d increased from narrow (34.8 ± 10.0 ms), to moderate (50.6 ± 8.4 ms), to prolonged (67.6 ± 18.3 ms). At AVDs 20 and 40 ms from optimal, CRI differed significantly (p < 0.001) between groups, with progressively higher CRI values as native QRS_d increased.

Electrical resynchronization with optimally programmed LV-only pacing was similar between patients with varying QRS_d, including patients with narrow QRS_d. The resynchronization window that corresponded with optimal electrical resynchronization decreased as native QRS_d decreased. This finding provides one potential explanation for the lack of significant benefit of CRT in narrow QRS_d patients in previous studies.