Abstract

AIMS: Following CRT, atrioventricular (AV) optimization is not routinely practised. To evaluate its clinical utility, we examined the effect of AV delay optimization on the prognostic biomarker NT-proBNP. METHODS AND RESULTS: We prospectively studied 72 patients (mean age 73 ± 12.5 years, 70.8% male, 55.6% ischaemic) undergoing iterative AV optimization. Patients were divided into those whose nominal setting appeared ideal and not changed (Group 1, n = 22) and those whose AV delay was optimized (Group 2, n = 50). All patients underwent NT-proBNP assessment prior to CRT, and pre- and a median 5 days post-optimization. Compared with Group 1, NT-proBNP fell significantly in Group 2 patients (median 474 pg/mL) following optimization (P = 0.00001). A significant change in filling pattern (defined as a change in AV delay >50 ms) was required in 30% of patients, and it was this subgroup that derived the greater reduction in NT-proBNP levels [-1407 pg/mL, interquartile range (IQR) -3042 to -346 pg/mL] compared with those requiring <50 ms AV delay change (-125 pg/mL, IQR -1038 to 6 pg/mL), P = 0.0011. The benefit of AV optimization was principally observed in reverse remodelling non-responders (median -2167 pg/mL, IQR -3042 to -305 pg/mL) and in patients with a pseudonormal or restrictive filling pattern (median -1407 pg/mL, IQR -2809 to -342 pg/mL), compared with those with more benign diastolic filling (median - 264 pg/mL, IQR -1038 to -21 pg/mL), P = 0.033. CONCLUSIONS: In one-third of patients, major filling pattern changes are achieved with AV optimization, associated with subsequent rapid falls in NT-proBNP. The greater the AV delay change, the larger the NT-proBNP fall, and non-responders and those with restrictive or pseudonormal filling despite CRT are most likely to benefit.