Prediction of long-term outcome of cardiac resynchronization therapy by acute pressure-volume loop measurements.

Aims Invasive assessment of acute haemodynamic response to biventricular pacing has been proposed as a tool to determine individual response and to optimize the effects of CRT. However, the long-term results of this approach have been poorly studied. The present study relates acute haemodynamic effects of CRT to long-term outcome. Methods and results Forty-one patients were analysed in the present study. During temporary biventricular pacing before implantation, acute changes in LV pump function were assessed by pressure–volume loop measurements and related to long-term response after CRT. In the study population [30 (71%) men, NYHA class 2.9 ± 0.4, EF 28 ± 7%, QRS 150 ± 25 ms], baseline mean stroke work (SW) and dP/dtmax were 4.6 ± 2.6 L ×mmHg and 874 ± 259 mmHg/s, respectively. During biventricular pacing, mean SW and dP/dtmax increased significantly by 43 ± 39% (+ 2.2 ± 2.4 L × mmHg, P < 0.001) and 13 ± 18% (+ 96 ± 136 mmHg/s, P < 0.001), respectively. In long-term responders (n = 29, 71%) compared with non-responders (n = 12, 29%), the acute increase in SW was significantly higher (+57 ±33% vs. + 10 ±30%, P < 0.001), whereas the acute increase in dP/dtmax was not significantly different between responders and non-responders (+ 15 ±18% vs. 6 ±15%, P = 0.139). Receiver operating characteristic (ROC) curve analysis indicated that SW was superior to dP/dtmax, QRS duration and LV dyssynchrony in prediction of response to CRT. A cut-off value for SW of 20% yielded a sensitivity of 90% and specificity of 75% to predict reverse remodelling at 6 months. Conclusion Invasive assessment of acute haemodynamics is a reliable tool to determine individual response to CRT. An acute increase in SW predicts long-term response to CRT with a higher accuracy than an acute increase in dP/dtmax, baseline QRS duration, and degree of LV mechanical dyssynchrony.