Effect of atrioventricular conduction prolongation on optimization of paced atrioventricular delay for biventricular pacing after cardiac surgery.

OBJECTIVES: Atrioventricular conduction prolongation (AVCP) in cardiac pacing is measurable and results primarily from delayed atrial conduction. Noninvasive methods for measuring atrial conduction are lacking. Accordingly, AVCP was used to estimate atrial conduction and investigate its role on the paced atrioventricular delay (pAVD) during biventricular pacing (BiVP) optimization. DESIGN: Retrospective analysis of data collected as part of a randomized controlled study of temporary BiVP after cardiopulmonary bypass. SETTING: Single-center study at university-affiliated tertiary care hospital. PARTICIPANTS: Cardiac surgical patients at risk of left ventricular failure after cardiopulmonary bypass. INTERVENTIONS: Temporary BiVP was optimized immediately after cardiopulmonary bypass. Vasoactive medication and fluid infusion rates were held constant during optimization. MEASUREMENTS AND MAIN RESULTS: For each patient the AVCP and the pAVD producing the optimum (highest) cardiac output (OptCO) and mean arterial pressure (OptMAP) were determined. Patients were stratified into long- and short-AVCP groups. Overall AVCP (mean ± standard deviation) was 64 ± 28 ms. For the short-AVCP group (<64 ms, n = 3), AVCP, OptCO, and OptMAP were 40 ± 11, 120 ± 0, and 150 ± 30 ms, respectively, and for the long-AVCP group (>64 ms, n = 4), these same parameters were 89 ± 10, 218 ± 44, and 218 ± 29 ms. OptCO and OptMAP were significantly less in the short-AVCP group (p = 0.015 and p = 0.029, respectively). CONCLUSIONS: AVCP varies widely after cardiopulmonary bypass, affecting optimum pAVD. Failure to correct for this can result in the selection of inappropriately short and potentially deleterious pAVDs, especially when nominal pAVD is used, causing BiVP to appear ineffective.

Primary endpoints of the biventricular pacing after cardiac surgery trial.

BACKGROUND: This study sought to determine whether optimized biventricular pacing increases cardiac index in patients at risk of left ventricular dysfunction after cardiopulmonary bypass. Procedures included coronary artery bypass, aortic or mitral surgery and combinations. This trial was approved by the Columbia University Institutional Review Board and was conducted under an Investigational Device Exemption. METHODS: Screening of 6,346 patients yielded 47 endpoints. With informed consent, 61 patients were randomized to pacing or control groups. Atrioventricular and interventricular delays were optimized 1 (phase I), 2 (phase II), and 12 to 24 hours (phase III) after bypass in all patients. Cardiac index was measured by thermal dilution in triplicate. A 2-sample t test assessed differences between groups and subgroups. RESULTS: Cardiac index was 12% higher (2.83±0.16 [standard error of the mean] vs 2.52±0.13 liters/minute/square meter) in the paced group, less than predicted and not statistically significant (p=0.14). However, when aortic and aortic-mitral surgery groups were combined, cardiac index increased 29% in the paced group (2.90±0.19, n=14) versus controls (2.24±0.15, n=11) (p=0.0138). Using a linear mixed effects model, t-test revealed that mean arterial pressure increased with pacing versus no pacing at all optimization points (phase I 79.2±1.7 vs 74.5±1.6 mm Hg, p=0.008; phase II 75.9±1.5 vs 73.6±1.8, p=0.006; phase III 81.9±2.8 vs 79.5±2.7, p=0.002). CONCLUSIONS: Cardiac index did not increase significantly overall but increased 29% after aortic valve surgery. Mean arterial pressure increased with pacing at 3 time points. Additional studies are needed to distinguish rate from resynchronization effects, emphasize atrioventricular delay optimization, and examine clinical benefits of temporary postoperative pacing.