The optimized V-V interval determined by interventricular conduction times versus invasive measurement by LVdP/dtMAX

We compared the calculated optimal V-V interval derived from intracardiac electrograms (IEGM) with the optimized V-V interval determined by invasive measurement of LVdP/dt(MAX).

Thirty-two patients with heart failure (six females, ages 68 +/- 7.8 years) had a CRT device implanted. After implantation of the atrial, right and a left ventricular lead, the optimal V-V interval was calculated using the QuickOpt formula (St. Jude Medical, Sylmar, CA, USA) applied to the respective IEGM recordings (V-V(IEGM)), and also determined by invasive measurement of LVdP/dt(MAX) (V-V(dP/dt)). The optimal V-V(IEGM) and V-V(dP/dt) intervals were 52.7 +/- 18 ms and 24.0 +/- 33 ms, respectively (P = 0.017), without correlation between the two. The baseline LVdP/dt(MAX) was 748 +/- 191 mmHg/s. The mean value of LVdP/dt(MAX) at invasive optimization was 947 +/- 198 mmHg/s, and at the calculated optimal V-V(IEGM) interval 920 +/- 191 mmHg/s (P < 0.0001). In spite of this significant difference, there was a good correlation between both methods (R = 0.991, P < 0.0001). However, a similarly good correlation existed between the maximum value of LVdP/dt(MAX) and LVdP/dt(MAX) at a fixed V-V interval of 0 ms (R = 0.993, P < 0.0001), or LVdP/dt(MAX) at a randomly selected V-V interval between 0 and +80 ms (R = 0.991, P < 0.0001).

Optimizing the V-V interval with the IEGM method does not yield better hemodynamic results than simultaneous BiV pacing. Although a good correlation between LVdP/dt(MAX) determined with V-V(IEGM) and V-V(dP/dt) can be constructed, there is no correlation with the optimal settings of V-V interval in the individual patient.