Europace Advance Access originally published online on January 29, 2008
Europace 2008 10(3):367-373; doi:10.1093/europace/eum287
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RESYNCHRONISATION THERAPY
Comparison of different approaches for optimization of atrioventricular and interventricular delay in biventricular pacing
Division of Cardiology, Kantonsspital Luzern, CH-6000 Luzern 16, Switzerland
Manuscript submitted 31 July 2007. Accepted after revision 10 December 2007.
* Corresponding author. Tel: +41 41 205 52 08; fax: +41 41 205 22 34. E-mail address: paul.erne{at}ksl.ch
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
Aims: It has been shown that optimizing atrioventricular (AV) and interventricular (VV) delay improves cardiac performance in patients with biventricular pacemakers. However, there is no standard method for optimization available yet. The aim of this study was to compare echocardiographic parameters—displacement imaging, A wave duration, and aortic velocity time integral (VTI)—and acoustic cardiography derived electromechanical activation time (EMAT) using different approaches of AV and VV delay optimization. We tested whether the initial optimization of the AV interval followed by VV optimization at that optimal AV interval or initial optimization of the VV interval followed by AV optimization at the determined optimal VV interval was accurate and consistent, and how this compared to testing every conceivable combination of AV and VV intervals available.
Methods and results: A group of 20 patients with biventricular pacemakers was included. Displacement imaging, A wave duration, and aortic VTI were determined at different combinations of AV (100, 150, 200, 250 ms) and VV (RV40, 0, LV40 ms) intervals. If AV duration was determined first, displacement imaging identified the best setting in 8/20, aortic VTI in 10/20, A duration in 13/20, and EMAT in 18/20 patients. With VV duration determined first, the best setting was more difficult to identify regardless of the method used. There was a poor agreement in optimal AV and VV delays of the different methods, and there was no single patient in whom all four methods yielded the same delay combination.
Conclusion: It is advisable to measure a full grid of AV and VV delays to identify optimal settings rather than optimizing one of the two delays first. Different techniques for delay optimization resulted in different optimal delay combinations.
Key Words: Acoustic cardiography, Biventricular pacing, Cardiac resynchronization therapy, Heart failure
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
Cardiac resynchronization therapy (CRT) is being used increasingly in patients with advanced heart failure. However, recent studies have reported the non-responder rate to biventricular pacing to be as high as 30–35%.1
,2 Factors that contribute to the high non-responder rate include poor patient selection, inappropriate lead positioning, and suboptimal device programming. Biventricular pacemaker settings have been optimized using a variety of techniques including echocardiography, surface ECG,3 and impedance cardiography.4,5 A new non-invasive technique, acoustic cardiography that is based upon the simultaneous recording of ECG and heart sounds, has also been shown to be effective for CRT optimization.6,7
The goals of atrioventricular (AV) delay optimization are to improve left ventricular (LV) filling and timing of contraction and to minimize mitral regurgitation.8 Various echocardiographic methods have been investigated for assessing AV delay optimization including those related to LV filling [the Ritter method, the iterative method,9 and to an index of cardiac output (aortic velocity time integral (VTI)10,11)]. Both systolic and diastolic parameters have been used to achieve LV optimization.
The goal of interventricular (VV) delay optimization is reduction of LV dyssynchrony to improve systolic performance.12 Estimating cardiac output by measuring the aortic VTI has been used to optimize VV delay.13 Echocardiographic displacement imaging detects left ventricular dyssynchrony by comparing the time difference to peak systolic contraction of the lateral and septal left ventricular walls.14
Acoustic cardiography (AudicorTM, Inovise Medical, Inc., Portland, OR, USA) has recently been developed to assess haemodynamics by measuring systolic time intervals and recording abnormal diastolic heart sounds using simultaneously recorded digital ECG and cardiac acoustical data. One of the systolic time intervals, electromechanical activation time (EMAT), has been shown to correlate well with invasive catheterization laboratory measurements of left ventricular function. EMAT, the time in ms from QRS onset to the mitral valve component of the first heart sound, has been used successfully to optimize AV delays.6,7
In practice, to manage the time required for the echocardiographic assessments, optimization is often done AV first (initial determination of AV interval followed by VV optimization at that optimal AV interval) or VV first (initial determination of VV interval followed by AV optimization at that optimal VV interval).
Since multiple parameters can be used for AV and VV optimization and multiple modalities are being used clinically, the goal of the present study was to compare four different parameters (acoustic cardiography, echocardiographic aortic VTI, A wave duration, and displacement imaging) in their consistency to identify the optimal delay setting using a full grid of settings. It was evaluated if determination of AV delay first or VV delay first is able to identify the same optimal AV and VV delay setting compared to testing every conceivable combination of AV and VV intervals available. We chose aortic VTI, since it relates to cardiac output and has been used for both AV and VV optimization,10,11,13 A wave duration due to higher reproducibility than other methods and correlation with left ventricular end-diastolic pressure, and displacement imaging as a measure of left ventricular dyssynchrony.14
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
Patients
Twenty patients, 18 men and 2 women, with biventricular pacemakers were enrolled in our study protocol. The study was approved by the local medical ethics committee and all patients provided written informed consent before they were enrolled. Average age was 64±10 years. The underlying cause of cardiac dysfunction was coronary heart disease in 8 patients, dilated cardiomyopathy in 11 patients, and giant cell myocarditis in 1 patient. The mean time from implantation of the biventricular pacemaker to study enrolment was 12 months (range 2–67 months). All the patients were on optimal medical therapy. All were receiving an ACE inhibitor or an angiotensin receptor blocker, 17 (85%) were on β-blocker therapy, 12 (60%) were getting spironolactone, and 13 (65%) were on a loop diuretic. Average NYHA class was 2.2 (range: 2–3) at study entry and all patients were in stable condition. None of the patients had atrial fibrillation.
Study protocol
A full echocardiographic examination was carried out in each patient with the biventricular pacemaker in standard programming (AV 120/150 ms, VV 0 ms) and an Audicor baseline examination was carried out. Table 1 summarizes baseline characteristics. Twelve AV and VV delay combinations were programmed (a grid of AV delays 100, 150, 200, 250 ms and VV delays RV40, 0, LV40) and aortic VTI, A wave duration, and displacement were measured. Forty-five delay combinations were tested with Audicor (a grid of AV delays at 100, 125, 150, 175, 200, 225, 250, 275, 300 ms and VV delays at RV40, RV20, 0, LV20, LV40 ms). However, to maintain comparability, the 12 delay combinations also used for the echocardiography parameters were analysed separately in this study. If intrinsic AV node conduction took place in more than 10% of heartbeats, no higher AV delays were tested.
|
Contour maps
Contour maps were generated from the full grids of AV and VV delays. The contour maps plot the AV and VV settings against the acoustic cardiographic or echocardiographic parameter being measured to graphically display the values of these diagnostic parameters. The contour maps make visualization of the maximum and minimum values easy as well as determination of the global maximum and minimum (single greatest or smallest value) vs. local maxima and minima. A contour map was considered to have a global maximum/minimum if all other local maxima/minima were <80% of the global value.
Echocardiography
All echocardiographic examinations were carried out by the same investigator (MZ) who was blinded to the Audicor results. All measurements were taken in eupnoea in left lateral supine position. The image quality of the echocardiograms was good in all examinations. A Toshiba Aplio 80 (Toshiba, Tokyo, Japan) with a 4 MHz transducer was used for Doppler echocardiography from the apical view according to the Guidelines of the American Society of Echocardiography. Displacement Imaging was used to measure the time difference between the maximal longitudinal contraction (in ms) of the septal and posterolateral mitral annulus in the apical four chamber view. Then, the echocardiographic probe was carefully aligned for measurement of the E- and A-wave velocity (cm/s), E deceleration time (ms), and A duration of the mitral inflow in ms (measured also at the tip of the mitral leaflets in the four chamber view), and aligned for the five chamber view to give accurate measurements of the aortic VTI in cm (measured in the left ventricular outflow tract approximately 1 cm below the aortic annulus with the PW-Doppler). Due to time constraints of the study, we measured dP/dt and the time from Q in the ECG to the beginning of mitral insufficiency in the CW-Doppler curve in those patients with mitral insufficiency at baseline examination, but not at all AV and VV delays.
Acoustic cardiography
The same investigator (ST) performed all acoustic cardiography examinations and was blinded to the echocardiographic results. After placing the patient in the supine position, an Audicor device was connected to record and trend acoustic cardiographic parameters using special sensors in the standard V3 and V4 positions that collect both digital sound and ECG. At each pacemaker setting, a 10 s Audicor recording was obtained and analysed by the computerized algorithm for measurement of various systolic time intervals. One of those systolic time intervals, the EMAT, is measured as the time from the onset of the Q wave to the mitral component of the S1. The S1 heart sound consists of acoustic energy from closure of both the mitral and tricuspid valves; however, the mitral valve component of S1 can be identified through its higher frequency component profile, so Audicor uses the first, most prominent high frequency component of the S1 as the marker for mitral valve closure.
The EMAT interval reflects the time required in ms for the LV to generate sufficient force to close the mitral valve. Since EMAT is a measure of contractility, the delay programming exhibiting shortest EMAT was considered optimal. The global optimal setting was determined based on EMAT from the full grid (45 settings) and to maintain consistency with the echocardiographic parameters also from the grid of 12 settings. In addition, a full recording with intrinsic conduction was done in all patients but those with high degree AV block (16/20 patients).
Reproducibility
We assessed reproducibility of EMAT, VTI, A duration, and Displacement Imaging by measuring 10 delay combinations twice in the same patient. We then assessed the correlation coefficient and the relative difference between the two measurements.
Statistical analysis
Data are presented as mean ± SD unless indicated otherwise. Frequencies were compared using Fisher’s exact test. Values of P < 0.05 were considered statistically significant. Pearson’s correlation was used for assessment of reproducibility and to assess correlation between intrinsic conduction and optimal AV delays.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
AV delays are presented as the AV delay active for intrinsic atrial activation. Values for paced atrium are 30 ms higher (sense compensation of –30 ms). Baseline measurements including echocardiographic and Audicor data with intrinsic conduction measurements, and data after optimization with Acoustic cardiography and Displacement Imaging are presented in Table 1.
‘AV first’ or ‘VV first’ implementation
The overall optimal AV and VV delays determined from the full AV–VV grids were compared to those that would have been determined using an ‘AV first’ approach and a ‘VV first’ approach (Table 2). The ‘AV first’ approach determines the optimal AV delay by setting VV = 0 ms and varying the AV delay then using that optimal AV delay while varying VV delay to find the optimal VV delay. The ‘VV first’ approach determines the optimal VV delay by setting AV = 150 ms (or closest setting) while varying the VV delay then using that optimal VV delay while varying AV to determine the optimal AV delay. Table 3 shows the number of cases the AV-first and VV-first method was able to identify the best setting for each modality. Aortic VTI and displacement imaging had the largest discrepancy between the overall optimal AV and VV delays compared to those determined using AV first or VV first. If using the AV first method, EMAT was able to identify 18/20 optimal settings, which was significantly more than aortic VTI and displacement imaging (P < 0.01).
|
|
Twelve measurements vs. 45 measurements grid
To estimate the effect of increasing the measurements number, we compared the best delay combination obtained with a grid of 12 and a grid of 45 EMAT measurements. There were only eight exact matches between the two protocols and mean difference between the optimal delays were 26 ± 39 ms and 14 ± 21 ms for AV and VV delay, respectively.
Contour maps
Contour maps for each patient for aortic VTI, A wave duration, displacement, and EMAT were constructed from the full AV–VV measurement grids and then inspected for the presence of global and local maxima (A wave duration, aortic VTI) as well as global and local minima (DI and EMAT). The ideal contour map would have a clearly defined global maximum or minimum rather than multiple local maxima or minima. Figure 1 contains contour maps for aortic VTI, A wave duration, DI, and EMAT for one patient. Most contour maps generated from the full AV–VV grid had local minima or maxima meaning that the ‘optimal’ delay determined from fewer tested settings would depend on those chosen to test. Aortic VTI had global maxima in 3 of 20 patients; A duration had global maxima in 7 of 20 patients; displacement imaging exhibited global minima in 9 of 20 patients (P = 0.05 compared to aortic VTI); EMAT had global minima in 10 of 20 patients (P = 0.02 compared to aortic VTI).
|
Correlation between the different optimal AV delays and with intrinsic PQ time
The best correlation was between the optimal AV delay determined by acoustic cardiography and A duration (r = 0.67). Intrinsic PQ time seems to influence optimal AV setting; we found significant correlations between intrinsic PQ time and optimal AV delay determined by acoustic cardiography (r = 0.68, P < 0.01), A duration (r = 0.74, P < 0.01) but not with optimal AV delay determined by Displacement Imaging (r = 0.28, P = 0.29) and aortic VTI (r = 0.21, P = 0.44).
Reproducibility
To compare reproducibility of the four different techniques, we measured 10 different delay settings twice at random order in the same patient. EMAT ranged from 94 to 186 ms for the 10 different delay combinations, A duration ranged from 82 to 153 ms, Displacement from 119 to 180 ms, and VTI from 11.6 to 15.3 cm. Thus, VTI expressed the least relative change between the different measurements. Correlation coefficient was 0.90, 0.85, 0.58, and 0.35 for EMAT, A duration, Displacement Imaging, and VTI, respectively. Variability between the two measurements was 9.9%, 8.5%, 8.6%, and 8.5% for EMAT, A duration, Displacement Imaging, and VTI, respectively.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
This study employs contour mapping to assess biventricular optimization methods including echocardiographic and acoustic cardiographic parameters. The contour maps for the echocardiographic and acoustic cardiography optimization methods did not have global maxima (VTI, A duration) or minima (displacement, EMAT) for most patients. This is clinically relevant because the lack of a global maximum/minimum means that without creating a full AV–VV grid only local maxima or minima will be located such as with an AV first or VV first approach (Figure 2). Vernooy et al.15
measured invasive left ventricular dP/dtmax and stroke volume in eight canine hearts was able to document global maxima for different AV delays and VV intervals. In another study, Whinnett et al.16 measured continuous blood pressure non-invasively and had global maxima in most of the contour maps for their patients when changing AV delays. The lack of a global maximum/minimum for some patients in this study explains why different optimal AV and VV settings are found with an AV first or VV first approach. The echocardiographic method in which AV first and VV first approaches were able to identify most of the overall maxima was optimization by A wave duration. The poor performance of aortic VTI has several reasons. First, it is technically difficult to place the sample exactly at the same place throughout the whole examination. Second, changes in aortic VTI are much smaller than changes of the three other parameters making identification of optimal intervals much more difficult.
|
Although intraobserver variability was almost identical for all four methods, EMAT, A duration, and Displacement Imaging performed better due to the larger change between the different delay settings. For example, in the patient chosen to measure reproducibility, EMAT varied between 94 and 186 ms whereas VTI values were between 11.6 and 15.3 cm making VTI much more vulnerable to a 10% intraobserver variability. A good parameter for AV- and VV-delay optimization is therefore a parameter which has small intraobserver variability and expresses large changes between the different delay settings. Such a parameter will probably produce a single optimal setting rather than multiple peaks in most patients.
Using the AV-first method for EMAT resulted in identical settings for 18 of the 20 patients to the full-grid optimal settings, and the VV-first method resulted in identical settings in 15 of 20 patients suggesting that a full grid may not be required. However, when the AV–VV grids for the full 45 settings were compared to the reduced grid of 12 settings, there were differences in the global optimal settings. No study that we are aware of has utilized a full grid of 45 settings for delay optimization with echocardiographic parameters. Based on these results for both echocardiographic and acoustic cardiographic methods, it is rational to create a full AV/VV map and select the best setting rather than identify the best setting for one delay first and then optimize the second.
The different optimization methods did not correlate well with each other in this study, similar to the findings of Kerlan et al.17 for aortic VTI and the mitral inflow method for AV optimization. In fact, in none of the 20 patients, all four methods yielded the same result. This may be due to assessment of different parts of ventricular function. While A wave duration clearly is a parameter of compliance in late diastolic filling mainly depending on AV time, displacement imaging measures time delay to maximal contraction between the septal and the lateral annulus indicating systolic dyssynchrony. Aortic VTI, and acoustic cardiography derived EMAT are parameters of global systolic heart function. However, we were able to show a significant correlation between optimal AV delays determined by acoustic cardiography and A duration and between AV delays determined by A duration and Displacement Imaging.
Acoustic cardiography offered some advantages compared to echocardiography. First, the advantage of analysing and averaging data obtained during 10 s (usually around 12 heart beats) and not just one heart beat. Second, the time required to perform an optimization using acoustic cardiography is shorter than that required for echocardiography. This becomes even more important if more delay combinations are tested. Optimization with acoustic cardiography (EMAT) was done in 
10 min for 12 settings by two operators. Optimization with echocardiography using aortic VTI and A duration was done by two operators (one performing the echocardiography, one changing pacemaker intervals). It again took 10 minutes to test twelve different settings. However, optimization with Displacement Imaging was much more time consuming due to the time for calculation needed and took around 20 minutes to test twelve settings.
Limitations
Although we have shown in a pilot study that optimization with acoustic cardiography leads to an improvement of echocardiography parameters and maximum oxygen consumption,7 we did not test the echocardiography parameters the same way. There was no parameter measured to compare the different optimal delay combinations, although no non-invasive gold standard exists. Forty-five delay combinations were tested with acoustic cardiography, while we tested only 12 delay combinations with echocardiography. We did not test RV triggered LV only pacing configuration that might have provided insight into fusion of intrinsic and paced activation.18
Summary
Optimization with different methods may lead to different optimal AV and VV delay settings. Regardless of the method chosen, it is advisable to test a full grid of different AV and VV combinations to identify the best setting rather than optimizing one of the two delays first.
|
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
Funding for this study was provided by the Swiss Heart Foundation.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
We gratefully thank Peter Bauer, PhD, and Patti Arand, PhD for their help in manuscript preparation.
Conflict of interest: none declared.
|
Footnotes |
|---|
Both authors contributed equally. 
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingAcknowledgementsReferences |
|---|
[1] Ansalone G, Giannantoni P, Ricci R, Trambaiolo P, Fedele F, Santini M. Biventricular pacing in heart failure: back to basics in the pathophysiology of left bundle block to reduce the number of nonresponders. Am J Cardiol (2003) 91:55F–61F.[CrossRef][Web of Science][Medline]
[2] Bax JJ, Abraham T, Barold SS, Breidthardt OA, Fung JW, Garrique S, et al. Cardiac resynchronization therapy: part 1—issues before device implantation. J Am Coll Cardiol (2005) 46:2153–67.
[3] Koglek W, Kranig W, Kowalski M, Kronski D, Brandl J, Oberbichler A. A simple method for AV-delay determination in dual chamber pacemakers. Herzchr Elektophys (2005) 11:244–53.
[4] Braun M, Schnabel A, Rauwolf T, Schulze M, Strasser RH. Impedance cardiography as a non-invasive technique for atrioventricular interval optimization in cardiac resynchronization therapy. J Interv Card Electrophysiol (2005) 13:223–9.[CrossRef][Web of Science][Medline]
[5] Heinroth KM, Elster M, Nuding S, Schlegel F, Christoph A, Carter J, et al. Impedance cardiography: a useful and reliable tool in optimization of cardiac resynchronization devices. Europace (2007) 9:744–50.
[6] Hasan A, Abraham WT, Quinn-Tate L, Brown L, Amkieh A. Optimization of cardiac resynchronization devices using acoustic cardiography: a comparison to echocardiography. Congest Heart Fail (2006) 12:25–31.[CrossRef][Medline]
[7] Toggweiler S, Zuber M, Kobza R, Roos M, Jamshidi P, Meier R, et al. Improved response to cardiac resynchronization therapy through optimization of atrioventricular and interventricular delays using acoustic cardiography. J Card Fail (2007) 13:637–42.[CrossRef][Medline]
[8] Gasparini M, Mantica M, Gallmberti P, La Marchesina U, Manglavacchi M, Faletra F, et al. Optimization of cardiac resynchronization therapy: technical aspects. Eur Heart J (2002) 4:D82–7.[CrossRef]
[9] Burri H, Lerch R. Utility of echocardiography for tailoring cardiac resynchronization therapy. Kardiovaskuläre Medizin (2006) 9:188–96.
[10] Kerlan JE, Sawhney NS, Waggoner AD, Chawla MK, Garhwal S, Osborn JL, et al. Prospective comparison of echocardiographic atrioventricular delay optimization methods for cardiac resynchronization therapy. Heart Rhythm (2006) 3:148–54.[CrossRef][Web of Science][Medline]
[11] Jansen AH, Bracke FA, van Dantzig JM, Meijer A, van der Voort PH, Aarnoudse W, et al. Correlation of echo-Doppler optimization of atrioventricular delay in cardiac resynchronization therapy with invasive hemodynamics in patients with heart failure secondary to ischemic or idiopathic dilated cardiomyopathy. Am J Cardiol (2006) 97:552–7.[CrossRef][Web of Science][Medline]
[12] Bax JJ, Abraham T, Barold SS, Breithardt OA, Fung JW, Garrique S, et al. Cardiac resynchronization therapy: part 2—issues during and after device implantation and unresolved question. J Am Coll Cardiol (2005) 46:2168–82.
[13] Parreira L, Santos JF, Madeira J, Mendes L, Seixo F, Caetano F, et al. Cardiac resynchronization therapy with sequential biventricular pacing: impact of echocardiography guided VV delay optimization on acute results. Rev Port Cardiol (2005) 24:1355–65.[Medline]
[14] Sade LE, Severyn DA, Kanzaki H, Dohi K, Gorcsan J. Second-generation tissue Doppler with angle-corrected color-coded wall displacement for quantitative assessment of regional left ventricular function. Am J Cardiol (2003) 92:554–650.[CrossRef][Web of Science][Medline]
[15] Vernooy K, Verbeek XAAM, Cornelussen RNM, Dijkman B, Crijns HJGM, Arts T, et al. Calculation of effective VV interval facilitates optimization of AV delay and VV interval in cardiac resynchronization therapy. Heart Rhythm (2007) 4:75–82.[CrossRef][Web of Science][Medline]
[16] Whinnett ZI, Davies JER, Willson K, Chow AW, Foale RA, Davies DW, et al. Determination of optimal atrioventricular delay for cardiac resynchronization therapy using acute non-invasive blood pressure. Europace (2006) 8:358–66.
[17] Kerlan JE, Sawhney NS, Waggoner AD, Shawla MK, Garhwal S, Osborn JL, et al. Prospective comparison of echocardigraphic atrioventricular delay optimization methods for cardiac resynchronization therapy. Heart Rhythm (2006) 3:148–54.[CrossRef][Web of Science][Medline]
[18] Stockburger M, Fateh-Moghadam S, Nitardy A, Langreck H, Haverkamp W, Dietz R. Optimization of cardiac resynchronization guided by Doppler echocardiography: haemodynamic improvement and intraindividual variability with different pacing configurations and atrioventricular delays. Europace (2006) 8:881–6.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||