Contractility surrogates derived from three-dimensional lead motion analysis and prediction of acute haemodynamic response to CRT

Background Patient-specific left ventricular (LV) lead optimisation strategies with immediate feedback on cardiac resynchronisation therapy (CRT) effectiveness are needed. The purpose of this study was to compare contractility surrogates derived from biventricular lead motion analysis to the peak positive time derivative of LV pressure (dP/dt(max) in patients undergoing CRT implantation.Methods Twenty-seven patients underwent CRT implantation with continuous haemodynamic monitoring. The right ventricular (RV) lead was placed in apex and a quadripolar LV lead was placed laterally. Biplane fluoroscopy cine films facilitated construction of three-dimensional RV-LV interlead distance waveforms at baseline and under biventricular pacing (BIVP) from which the following contractility surrogates were derived; fractional shortening (FS), time to peak systolic contraction and peak shortening of the interlead distance (negative slope). Acute haemodynarnic CRT response was defined as LV Delta dP/dt(max ) >= 10 %.Results We observed a mean increase in dP/dt(max) under BIVP (899 +/- 205 mm Hg/s vs 777 +/- 180 mm Hg/s, p<0.001). Based on Delta dP/dt(max), 18 patients were classified as acute CRT responders and nine as non-responders (23.3%1 +/- 10.6% vs 1.91 +/- 5.3%, p<0.001). The baseline RV-LV interlead distance was associated with echocardiographic LV dimensions (end diastole: R=0.61, p=0.001 and end systole: R=0.54, p=0.004). However, none of the contractility surrogates could discriminate between the acute CRT responders and non-responders (Delta FS: -2.5 +/- 2.6% vs - 2.0 +/- 3.1%, p=0.50; Delta time to peak systolic contraction: -9.7 +/- 18.1% vs -10.8 +/- 15.1%, p=0.43 and Delta peak negative slope: -8.7 +/- 45.9% vs 12.5 +/- 54.8 %, p=0.09).Conclusion The baseline RV-LV interlead distance was associated with echocardiographic LV dimensions. In CRT recipients, contractility surrogates derived from the RV-LV interlead distance waveform could not discriminate between acute haemodynamic responders and non-responders