(BQ) Part 2 book Cardiac resynchronization therapy presents the following contents: Optimization of the interventricular (V-V) interval during cardiac resynchronization therapy, complications of cardiac resynchronization therapy, asynchrony in coronary artery disease, cardiac resynchronization therapy in right bundle branch block,...
9781841846378-Ch12 6/14/07 10:08 AM Page 165 12 Optimization of the interventricular (V–V) interval during cardiac resynchronization therapy S Serge Barold, Arzu Ilercil, Stéphane Garrigue, and Bengt Herweg Programmability of the interventricular interval • Pathophysiologic basis for programming the V–V interval • Clinical studies of V–V interval programming • General considerations • Effect of V–V timing on the ECG of biventricular pacemakers • Automatic device-based optimization of the V–V delay PROGRAMMABILITY OF THE INTERVENTRICULAR INTERVAL PATHOPHYSIOLOGIC BASIS FOR PROGRAMMING THE V–V INTERVAL The methods for atrioventricular (AV) optimization in patients receiving cardiac resynchronization therapy (CRT) are almost universally used for programming the optimal interventricular (V–V) delay.1–6 Conventional M-mode echocardiography for the measurement of left ventricular (LV) dyssynchrony using septal-to-posterior wall motion delay may be unreliable and poorly reproducible.7 Determination of the extent of residual LV dyssynchrony after V–V programming requires more sophisticated echocardiographic techniques such as tissue Doppler techniques (peak velocity time difference, delayed longitudinal contraction score, etc.), three-dimensional (3D) echocardiography, and automatic endocardial border detection.8–12 Contemporary biventricular devices permit programming of the V–V interval usually in steps from +80 ms (LV first) to −80 ms (right ventricle (RV) first) to optimize LV hemodynamics This design was the result of cogent pathophysiologic considerations that simultaneous activation of the two ventricles for CRT was illogical.13 Perego et al13 advanced arguments that the best mechanical efficiency in CRT is not necessarily achieved by simultaneous pacing of the two ventricles (hence the importance of programmability of the V–V interval) (Figure 12.1): In normal hearts, activation of the two ventricles does not occur simultaneously, i.e., epicardial RV depolarization starts a few milliseconds earlier than LV depolarisation.14,15 In CRT, epicardial LV pacing delays transmission of activation that is normally supposed to reach the subendocardial conduction system before it spreads to the remaining ventricle In advanced cardiomyopathy, RV-to-LV interactions can be different from those in normal hearts Myocardial disease is associated with different locations and sizes of scars, and heterogeneity of conduction disturbances The baseline ventricular conduction defect differs considerably from case to case, especially in patients with a QRS duration >150 ms.16 Theoretically, slow conduction in 9781841846378-Ch12 6/14/07 10:08 AM Page 166 166 CARDIAC RESYNCHRONIZATION THERAPY possibly with right bundle branch activation alter QRS configuration and hemodynamics LVp LVp Delayed conduction Delayed conduction Time Normal conduction RVp SIMULTANEOUS PACING Normal conduction RVp Programmable V–V delay SEQUENTIAL PACING Figure 12.1 Diagrammatic representation of left ventricular (LV) conduction delay interfering with synchronous activation of the two ventricles at the broken horizontal line Programmability of the interventricular (V–V) interval permits pre-activation of the LV to compensate for the LV conduction delay In this way, both ventricles are activated synchronously at the broken horizontal line LVp, LV pacing event; RVp, right ventricular pacing event the presence of scar tissue in ischemic cardiomyopathy would necessitate more LV pre-excitation Conduction delay may be caused not only by isolated left bundle branch block (LBBB), but also by more global anisotropic disturbances of the conduction system and/or myocardial scars, latency of LV stimulation, and delayed global depolarization.17–20 Despite similar QRS morphology, congestive heart failure (CHF) patients with LBBB, and LV dyssynchrony exhibit different locations and patterns of dyssynchrony.21 The ventricular leads (particularly the LV leads) are placed in quite different anatomic positions, depending on the operator’s choice and coronary sinus anatomy, producing paced ventricular activation patterns that differ from patient to patient V–V programmability may compensate for less than optimal LV lead position by tailoring ventricular timing to correct for individual heterogeneous ventricular activation patterns commonly found in patients with LV dysfunction and CHF The presence and varying degree of fusion with the spontaneous QRS complex and On the basis of the above arguments, it is therefore not surprising that V–V programmability in the reported studies has shown a heterogeneous response, with great variability of the optimal V–V delay from patient to patient, so that adjustment of the V–V delay, like the AV delay, must be individualized (Figures 12.1 and 12.2) In addition, assessment of the role of V–V programmability is compounded by the varied cut-off QRS duration for inclusion in the various studies, the different testing procedures to determine the optimum V–V delay, and whether AV delay optimization was performed before testing the V–V response CLINICAL STUDIES OF V–V INTERVAL PROGRAMMING Although V-V programmability produces a rather limited improvement in stroke volume, the response is important in patients with a less than desirable response to CRT It is presently unknown whether AV and/or V–V interval optimization can actually decrease the percentage of non-responders to CRT Sogaard et al21 performed one of the first studies evaluating the role of V–V delay in CRT patients, and convincingly demonstrated that the site and degree of mechanical asynchrony can vary from patient to patient and are influenced by the underlying etiology of disease, whether ischemic or non-ischemic They defined a new parameter that they called the extent of delayed LV longitudinal contraction (DLC) (Figure 12.3) This is calculated using tissue Doppler imaging (TDI) coupled with strain rate analysis A segment was considered to have DLC if the strain rate analysis demonstrated motion reflecting true contraction and if the end of the segmental contraction occurred after aortic valve closure Sogaard et al21 found that the extent of myocardium with DLC predicted improvement of LV systolic performance and reversion of LV remodeling during short- and long-term CRT Their observations indicated that DLC represented mechanical LV asynchrony and thus a contractile reserve, which could be recruited by CRT (Figure 12.3) 9781841846378-Ch12 6/14/07 10:08 AM Page 167 OPTIMIZATION OF THE INTERVENTRICULAR INTERVAL DURING CRT 167 0.785 × (diameterLVOT)2 × VTILVOT = SV V–V −80 ms delay −40 ms Simultaneous RV pre-excitation +40 ms +80 ms LV pre-excitation LVOT VTI V–V time corresponding to greatest stroke volume Figure 12.2 Interventricular V–V interval delay using left ventricular outflow tract (LVOT) measurements of blood flow velocities for estimation of stroke volume (SV) SV is exponentially related to the LVOT diameter and directly to the velocity–time integral (VTI) of the LVOT Variation of the V–V interval affects the SV, as evidenced by varying VTI measurements that can serve as surrogate markers for resynchronization The optimal V–V interval in this example is derived from pacing the right ventricle (RV) 40 ms before the left ventricle (LV) The optimal AV delay becomes equal to optimal AS-LVP minus the 40 ms V–V interval LVP, monochamber LV pacing (Reproduced from Gassis S, Leon AR Cardiac resynchronization therapy: strategies for device programming, troubleshooting and follow-up J Interv Card Electrophysiol 2005;13:209–22.) However, the location of myocardium displaying DLC is variable in patients with heart failure and ventricular conduction disturbances It was hypothesized that individually tailored preactivation of myocardium displaying DLCs could further improve the overall response to CRT Sogaard et al,21 using Doppler imaging techniques, studied 21 patients with LBBB, QRS > 130 ms, and New York Heart Association (NYHA) functional class III or IV heart failure, specifically before and after CRT (Figure 12.4) Post-implantation studies were performed during simultaneous CRT and at 12, 20, 40, 60, and 80 ms V–V delay intervals, with either LV or RV preexcitation The study population consisted of 11 patients with ischemic cardiomyopathy and patients with idiopathic dilated cardiomyopathy As noted in prior studies, DLC in patients with idiopathic dilated cardiomyopathy was identified in the lateral and posterior LV walls In contrast, ischemic cardiomyopathy exhibited DLC more frequently in the septal and inferior walls Echocardiographic parameters improved during sequential CRT, with LV pre-activation being superior in patients and RV pre-activation being superior in 11 patients (Figure 12.4) Compared with simultaneous CRT, tailored sequential CRT reduced the extent of segments with DLC in the base from 23 ± 13% to 11 ± 7% (p130 у150 29 22 2005 2005 20 207 BiV0, 359 sequential у130 у130 30 31 2006 2006 23 86 >120 >150 Invasive LV dP/dtmax Sequential pacing 41% pts, with only RV1 pt Others BiV0 equivalent Radionuclide angiography (LVEF) Echocardiography MPI Echo Doppler determination of cardiac output LVOT VTI Echo Doppler determination of stroke volume LV1 45%, BiV0 33%, RV1 22% LV1 48%, RV1 48%, BiV0 4% LV1 best in most pts, RV1 best in pts LV1 12, RV1 5, BiV0 pts At months: LV1 58%, BiV0 19%, RV1 23%, LV1 60, BiV0 22%, RV1 18% LV1 36%, RV1 35%, BiV0 29% Aortic VTI Echo Doppler determination of stroke volume 3D, 3-dimensional; AF, atrial fibrillation; BiV0, simultaneous biventricular pacing; LBBB, left bundle branch block; LV, left ventricle; LV1, LV pre-activation; LVEF, LV ejection fraction; LVOT, LV outflow tract; MPI, myocardial performance index PM, pacemaker; pts, patients; RBBB, right bundle branch block; RV, right ventricle; RV1, RV pre-activation; SR, sinus rhythm; TDI, tissue Doppler imaging; VTI, velocity–time integral a The results indicate the distribution of the optimal V–V delay according to its corresponding pacing mode: LV1 , RV1, and BiV0 in terms of the number of patients or percentage All patients were in sinus rhythm unless indicated otherwise (AF) Long-term stability of the optimal V–V interval and clinical response The optimal V–V delay may change with the passage of time, and individual changes cannot be accurately predicted Detailed, regular re-evaluations and reprogramming of optimal parameters seem appropriate Boriani et al31 reported disappointing results at the 6-month follow-up after V–V optimization They selected patients at random and compared the results of CRT with simultaneous biventricular pacing (n = 23) versus V–V optimized devices (n = 72) after a follow-up of months There were no differences in symptoms, quality of life, or functional capacity between the two groups These results are difficult to explain, but they may be related to the selection of sicker patients (QRS у150 ms), the lack of AV optimization after programming the V–V interval, a change in the optimal V–V interval after months, or progression of disease In this respect, O’Donnell et al33 studied 40 recipients of CRT devices Optimized V–V delays were determined according to echocardiographic criteria There was a trend toward reduction in the LV predominance of the optimal 9781841846378-Ch12 6/14/07 10:09 AM Page 173 OPTIMIZATION OF THE INTERVENTRICULAR INTERVAL DURING CRT 173 V–V delay during follow-up The mean optimal V–V delay at implantation was 22 ms (range −12 to +32 ms) with the LV activated first, versus 12 ms (range −16 to +32 ms) at months These observations are partially supported by the data of Mortensen et al,25 who found that the optimal V–V interval changed in 56% of CRT patients at the 3-month follow-up V–V interval optimization on exercise A recent study assessed the impact of sequential biventricular pacing during exercise.30 Simultaneous biventricular pacing was optimal during exercise in only about 25% of patients (Figure 12.7) Most of the improvement was observed with short V–V delays, ranging from 12 to 20 ms Optimized sequential biventricular pacing offered substantial additional benefit when considering the aortic VTI and mitral regurgitation Differences between resting and exercise optimization were observed in more than half of the patients With future technological advances, separate automatic programming between resting and exercise for V–V delay may become possible by means of sensors or other ways to control hemodynamics at rest and with activity Recent data from the same group suggest that the degree of LV dyssynchrony varies with exercise and may diminish in some patients Percentage of patients 30 25 Simultaneous 20 RV20 RV12 LV12 15 10 LV20 LV40 EFFECT OF V–V TIMING ON THE ECG OF BIVENTRICULAR PACEMAKERS The electrocardiographic (ECG) consequences of temporally different RV and LV activation with programmable V–V timing in the latest biventricular devices have not yet been studied in detail In the absence of anodal stimulation, increasing the V–V interval gradually to 80 ms (LV first) will progressively increase the duration of the paced QRS complex and alter its morphology, with a larger R wave in lead V1, indicating more dominant LV depolarization.34 The varying QRS configuration in lead V1 with different V–V intervals has not been correlated with the hemodynamic response Consequently, at this, juncture it is unwise to attempt programming the optimal V–V interval according to the height of the paced R wave in lead V1 Anodal stimulation RV anodal stimulation during biventricular pacing interferes with a programmed V–V delay (often programmed with the LV preceding the RV) aimed at optimizing cardiac resynchronization This interference occurs because RV anodal capture causes simultaneous RV and LV activation (the V–V interval becomes zero) In the presence of anodal stimulation, the ECG morphology and its duration will not change if the device is programmed with V–V intervals of 80, 60, and 40 ms (LV before RV) The delayed RV cathodal output (80, 60, and 40 ms) then falls in the myocardial refractory period initiated by the preceding anodal stimulation At V–V intervals р 20 ms, the paced QRS may change because the short LV–RV interval prevents propagation of activation from the site of RV anodal capture in time to render the cathodal site refractory.34 Thus, the cathode also captures the RV and contributes to RV depolarization, which then takes place from two sites: RV anode and RV cathode.34 Rest Exercise Figure 12.7 Optimal V–V delay at rest and during exercise RV20, RV lead pre-excitation 20 ms, etc.; LV12, LV lead preexcitation 12 ms, etc (Reproduced from Bordachar P et al Am J Cardiol 2006;97:1622–5.30) AUTOMATIC DEVICE-BASED OPTIMIZATION OF THE V–V DELAY St Jude Medical have recently introduced a method whereby the programmer itself can 9781841846378-Ch12 6/14/07 10:10 AM Page 174 174 CARDIAC RESYNCHRONIZATION THERAPY 60 y = 0.9841x + 1.4614 50 Max aortic VTI r = 97.69% 40 30 20 10 n = 61 patients 0 10 20 30 40 50 60 Aortic VTI at IEGM VV (cm) Figure 12.8 Comparison of the aortic velocity–time integral with the corresponding value obtained from analysis of intracardiac electrograms (IEGM) (Reprinted from Heart Rhythm; 3(Suppl.)) Meine M, Min X, Paris M, et al An intracardiac EGM method for VV optimization during cardiac resynchronization (Abstract) Pages S63–S64 (2006).36 determine and then program the V–V delay automatically.35 This design was based on a study involving 61 patients who received a St Jude EPIC HF device, which used the ventricular electrogram (IEGM) to obtain the optimal V–V interval.36 Optimal V–V delays based on the IEGM algorithm were compared with the optimal V–V interval obtained by the maximum aortic VTI over seven V–V delays (20, 40, and 80 ms), with both RV and LV leads pre-activated and simultaneous biventricular pacing (Figure 12.8) The maximum aortic VTI (22.1 ± 8.2 cm) was equivalent to the IEGM aortic VTI values (20.9 ± 8.3 cm) (concordance r = 0.98 and a 95% confidence lower limit of 97%; p 120 >130 у150 29 22 20 05 20 05 20 20 7 BiV0, 359 sequential у130 у130 30 31 20 06 20 06 23 86 > 120 >150 Invasive LV dP/dtmax Sequential... 9781841846378-Ch 12 6/14/07 10:09 AM Page 1 72 1 72 CARDIAC RESYNCHRONIZATION THERAPY Table 12. 1 Studies of sequential biventricular pacing Ref Year No of pts QRS (ms) Parameter Resultsa 21 13 23 20 02 2003 20 04... Cardiol 20 05;46 :22 98–304 .22 ) p