Beneficial Effects of Biatrial Pacing on Cardiac Function in Patients With

Circ J 2005; 69: 831–836

Beneﬁcial Effects of Biatrial Pacing on Cardiac Function in Patients With Bradycardia–Tachycardia Syndrome

Katsumi Matsumoto, MD; Toshiyuki Ishikawa, MD; Shinichi Sumita, MD; Kohei Matsushita, MD; Noriko Kawasaki, MD; Tsukasa Kobayashi, MD; Kazuaki Uchino, MD; Kazuo Kimura, MD; Satoshi Umemura, MD

Background Biatrial (BiA) pacing prevents atrial fibrillation. By an unknown mechanism. The purpose of this study was to use Doppler echocardiography to evaluate the hemodynamic effects during BiA pacing. Methods and Results The subjects were 7 patients with bradycardia–tachycardia syndrome with an implanted pacemaker. Atrial pacing sites were the right atrial appendage (RAA) and coronary sinus. P wave duration during BiA pacing (123±16ms) was significantly shorter than during either RAA pacing (167±19ms, p<0.05) or sinus rhythm (148±12ms, p<0.05). Doppler echocardiography revealed a greater cardiac output during BiA pacing than during RAA pacing (4.1±1.1 vs 3.5±0.7L/min, p=0.042). The Doppler waveform of transmitral flow indicated that the left ventricular contraction interrupted the atrial filling wave during RAA pacing. The interval between the end of the atrial filling wave of transmitral flow and the mitral valvular closing sound was significantly increased by BiA pacing compared with RAA pacing (56±65 vs 40±57ms, p=0.047). Conclusion Cardiac hemodynamics were improved by BiA pacing and reduction of left atrial load may be one of the mechanisms. (Circ J 2005; 69: 831–836) Key Words: Atrial fibrillation; Biatrial pacing; Cardiac function; Intra-atrial conduction delay; Sick sinus syndrome

trial fibrillation (AF) is a common arrhythmia1 that is (SSS) with a class I indication for an implanted pacemaker sometimes resistant to anti-arrhythmic therapy.2–5 and frequent recurrences of paroxysmal AF. Six of them A Biatrial (BiA) pacing is a preventive therapy6–8 also had common atrial flutter for which they had under- that is thought to act via an improvement in the intra-atrial gone radiofrequency catheter ablation to the isthmus conduction delay;9–11 however, its actual mechanism is between the tricuspid valve and inferior vena cava and had unclear. Some investigators reported that BiA pacing did been treated with a Na-channel blocker or another anti- not change the cardiac hemodynamics,12,13 but Doi et al arrhythmic agent. Left ventricular diastolic dimension reported that BiA pacing yielded hemodynamic benefits (LVDd) and left ventricular ejection fraction (LVEF) were compared with high right atrial pacing.14 We hypothesized within normal limits (LVDd=44.3±4.7mm, LVEF=71.0± that BiA pacing might improve cardiac hemodynamics 8.7%), but the left atrial dimension was slightly large compared with right atrial appendage (RAA) pacing be- (41.8±5.6mm). All patients underwent implantation of a cause of differences in the relationship between the con- DDD pacemaker (Medtronic THERA® DR 7968i). BiA traction of the left atrium and left ventricle during each pacing was performed with an additional lead: one lead pacing mode. Left atrial contraction is delayed during RAA (Medtronic 4592–53) was positioned in the RAA and pacing because of the inter- or intra-atrial conduction de- another (Medtronic 2188–58) was positioned within the lay, so we inferred that this delay has an adverse effect on coronary sinus (CS). Both atrial leads were connected to cardiac function. The aim of our study was to use Doppler the atrial port of the pacemaker by a Y-adapter, with the CS echocardiography to evaluate the efficacy of BiA pacing lead acting as the anode and the RAA lead as the cathode. against the deleterious effect of RAA pacing. The CS leads were positioned from the lateral to the posterolateral portion of the CS. Bipolar pacing achieved BiA pacing, while unipolar pacing achieved RAA pacing. Methods The ventricular lead (Medtronic CAPSURE® SP 4023–58) Subjects was positioned in the right ventricular apex and connected The subjects were 7 patients (mean age, 68±6.6 years, 4 to the ventricular port of the pacemaker (Fig1). BiA pacing male; Table1). None had ischemic heart disease or idiopa- was approved by the Ethical Committee of Yokohama City thic cardiomyopathy, but all had sick sinus syndrome University Hospital and informed consent was obtained from each participant before enrollment in the study. (Received January 5, 2005; revised manuscript received April 25, 2005; accepted May 2, 2005) Study Protocol Second Department of Internal Medicine, Yokohama City University Doppler echocardiography was performed with patients School of Medicine, Yokohama, Japan Mailing address: Katsumi Matsumoto, MD, Second Department of resting in the left lateral position with electrocardiogram Internal Medicine, Yokohama City University School of Medicine, (ECG) and heart sound monitoring, after 10min of BiA 3-9 Fukuura, Kanazawa-ku, Yokohama 236-0004, Japan. E-mail: pacing and RAA pacing. Measurement of the Doppler birth-45 @fukuhp.yokohama-cu.ac.jp waveform of the transmitral flow was performed using an

Circulation Journal Vol.69, July 2005 832 MATSUMOTO K et al.

Table 1 Baseline Characteristics of the Patients

Patient Age (years), Mode rate C-AFL LVEF LVDd LAD Anti-arrhythmic agent no. sex (beats/min) RF-CA (%) (mm) (mm) 1 63M AAI70 Y Pilsicainide, sotalol, digoxin 61.1 42.4 34.9 2 61M AAI60 Y Pilsicainide 74.6 42.7 42.7 3 75F AAI70 N None 66.8 42.1 42.7 4 64M AAI70 Y Pilsicainide, sotalol, atenolol, diltiazem 75.4 44.5 40.3 5 70M AAI70 Y Pilsicainide, digoxin 59.0 54.6 39.2 6 78F AAI70 Y Pilsicainide, atenolol 81.0 43.0 40.0 7 64F DDD85 Y Pilsicainide, sotalol, varapamil, metildigoxin 79.1 40.8 52.9 Mean 67.9±6.6 years 71.0 44.3 41.8

C-AFL, common atrial flutter; RF-CA, radiofrequency catheter ablation; LVEF, left ventricular ejection fraction; LVDd, left ventricular diastolic dimension; LAD, left atrial dimension.

Fig1. Chest X-ray showing biatrial pacemaker (Case 3). RAA, right atrial appendage; CS, coronary sinus; RVA, right ventricular apex. apical 2-chamber view with the sample volume between transducer. Measurement parameters were as follows: in the mitral leaflet tips, and that of aortic valve flow was the apical 2-chamber view, peak velocity of the early filling performed using an apical 3-chamber view with the sample wave of the transmitral flow (E, cm/s), peak velocity of the volume between the aortic leaflet tips. The echocardio- atrial filling wave of the transmitral flow (A, cm/s), E/A grapher was unaware of the pacing mode. Doppler was ratio, deceleration time (DT, ms), interval between pacing recorded at a sweep speed of 100mm/s, and data based on spike and A, end of atrial filling wave (A end), mitral valve the mean of 3 typical waveforms were analyzed later. In 6 closure sound (S-A, ms; S-A end, ms; S-I, ms), and interval patients with normal atrioventricular (AV) conduction, the between the A end and mitral valve closure sound (A end-I, pacing mode was AAI 60 or 70beats/min. In 1 patient with ms). In the apical 3-chamber view, stroke volume (SV, ml) advanced AV conduction block, the pacing mode was DDD and ejection time (ET, s) were measured, and cardiac output 85beats/min and the optimal AV delay was determined by (CO, L/min) was calculated from the aortic flow velocity a previously reported method.15 Measurement parameters integral. were indicated afterward. All measurements were performed 5 times, and the average values determined. Statistical Analysis Values are given as mean±standard deviation. Compari- Electrocardiography son between BiA and RAA pacing was performed by Surface 12-lead ECG was recorded in all patients during paired t-test, and among sinus rhythm, BiA and RAA with sinus rhythm, RAA pacing and BiA pacing. The duration respect to P duration was performed by Tukey-Kramer of the P wave was measured, and the interval between the post-hoc test after ANOVA. atrial spike of the pacemaker and QRS complex (S-Q interval) was determined. The longest P wave duration and S-Q interval were compared. Results Electrocardiography Echocardiographic Measurements The P wave duration during BiA pacing (123±16ms) Doppler echocardiographic studies were performed with was significantly shorter than during either RAA pacing a commercially available instrument (Hewlett-Packard (167±19ms, p<0.05) or sinus rhythm (148±12ms, p<0.05) Sonos 2500, Andover, MA, USA) equipped with a 2.5MHz (Fig 2) and the S-Q interval during BiA pacing (191±

Circulation Journal Vol.69, July 2005 Biatrial Pacing and Cardiac Function 833

Fig 2. Comparison of P duration. P duration during biatrial (BiA) pacing was signiﬁcantly shorter than that during right atrial appendage (RAA) pacing and that during sinus rhythm (SR).

Fig3. Doppler waveform of mitral valve flow in Case 3. A end-I was 20ms during biatrial (BiA) pacing (a), and was 0ms during right atrial appendage (RAA) pacing, and the A wave was interrupted by ventricular contraction (arrow, b). E, early filling wave; A, atrial filling wave; S1, first heart sound.

23ms) was significantly shorter than during RAA pacing Doppler studies. There were significant differences in the (226±34ms, p=0.0023). measurements of aortic valve flow: during BiA pacing, SV was greater (56±14 vs 50±10ml, p=0.01), ET was longer Echocardiographic Measurements (304±31 vs 293±35ms, p=0.044), and CO was greater (4.1± There was no case of valvular disease in the pulsed- 1.1 vs 3.5±0.7L/min, p=0.042) than during RAA pacing.

Circulation Journal Vol.69, July 2005 834 MATSUMOTO K et al.

Fig4. Schema of biatrial (BiA) pacing (A) and right atrial appendage (RAA) pacing (B). PQ interval on the right side during BiA pacing is shorter than that during RAA pacing. PQ interval on the left side during BiA pacing is longer than that during RAA pacing.

Table 2 Electrocardiographic and Echocardiographic Measurements BiA pacing showed only a straight line with A end-I of BiA RAA p value 20ms, whereas during RAA pacing it was curved from the peak of the atrial filling wave to the end of it and there was S-Q (ms) 191±23 226±34 0.0023 no interval of A end-I. This warped line was caused by the SV (ml) 56±14 50±10 0.01 left ventricular contraction before the end of left atrial ET (ms) 304±31 293±35 0.044 CO (L/min) 4.1±1.1 3.5±0.7 0.042 systole and its shape indicates that the left ventricular E (cm/s) 65±32 58±38 0.014 contraction interrupted the left atrial contraction. However, A (cm/s) 67±18 62±17 0.009 there was no meaningful change in the Doppler waveforms E/A 1.05±0.70 1.03±0.81 NS of mitral valve flow during both pacing modes in the other DT (ms) 180±52 195±83 NS 4 patients, included the one with advanced AV conduction S-A (ms) 167±6 215±13 0.0035 S-A end (ms) 241±18 286±17 0.0001 block. Overall, A end-I during BiA pacing (56±65ms) was A end-I (ms) 56±65 40±57 0.047 significantly longer than during RAA pacing (40±57ms, S-I (ms) 297±74 326±66 0.033 p=0.047). The intervals from the pacing spike to A (S-A), A end (S- BiA, biatrial pacing; RAA, right atrial appendage pacing; S-Q, interval between atrial spike of pacemaker and QRS complex; SV, stroke volume; ET, A end) and I (S-I) were all shorter during BiA pacing than ejection time; CO, cardiac output; E, peak early filling velocity; A, peak during RAA pacing (S-A: 167±6 vs 215±13ms, p=0.0035; atrial filling wave velocity; DT, deceleration time; S-A, interval between S-A end: 241±18 vs 286±17ms, p=0.0001; S-I: 297±74 vs pacing spike and peak atrial filling wave velocity; S-A end, interval be- 326±66ms, p=0.033). tween pacing spike and end of atrial filling wave; A end-I, interval between Both E and A during BiA pacing were significantly end of atrial filling wave and mitral valve closure sound; S-I, interval between pacing spike and mitral valve closure sound. higher than during RAA pacing (65±32 vs 58±38 ms, p=0.014; 67±18 vs 62±17ms, p=0.009). However, there was no significant difference in the E/A ratio between BiA There were different Doppler waveforms of mitral valve and RAA pacing (1.05±0.70 vs 1.03±0.81, NS) or in DT flow in 3 patients during both pacing modes and 2 of them (180±52 vs 195±83ms, NS). were receiving a Na-channel blocker. Fig3 is an example of the different Doppler waveforms. The waveform during

Circulation Journal Vol.69, July 2005 Biatrial Pacing and Cardiac Function 835

that BiA pacing produced an improvement in CO compared Discussion with high right atrial pacing because of atrial resynchro- Atrial fibrillation is a common arrhythmia, especially nization.14 They used a temporary DDD pacing mode, with in patients with SSS.16 Many patients with AF suffer from 100ms or 150ms AV delay and without antiarrhythmic severe symptoms, such as palpitations,17 and AF is associ- agents. We used AAI pacing or DDD pacing with optimal ated with a high incidence of cerebral infarction18 and in- AV delay using an implanted pacemaker and antiarrhyth- creased cardiovascular mortality.19 AF is caused by an mic agents and we consider that our study achieved a sig- abnormal atrial substrate that has an inter-atrial conduction nificant clinical result because the P wave duration during delay20 and short atrial refractory period with increased sinus rhythm was larger than in the 3 previous studies12–14 dispersion. BiA pacing, a type of multi-site atrial pacing, and the intra-atrial conduction delay was prolonged. prevents the re-occurrence of AF,6–8 by improving both CO was greater during BiA pacing because of an im- the electrophysiological mechanism and hemodynamic provement in both the timing of inter-atrial conduction and changes. BiA pacing is assumed to suppress the dispersion the relationship between the contraction of the left atrium of atrial contraction, and to improve the inter- or intra-atrial and ventricle. The deleterious change in cardiac function conduction delay,9–11 but the hemodynamic mechanism during RAA pacing might be caused by a rise in left atrial remains unclear. Some investigators have reported that BiA and left ventricular end-diastolic pressures because the left pacing had no benefit on hemodynamics,12,13 whereas Doi ventricular contraction starts before termination of the et al reported that BiA pacing was beneficial compared transmitral flow of left atrial contraction. Not only was with high right atrial pacing.14 aortic flow greater but ET was also prolonged by BiA We hypothesized that BiA pacing might improve cardiac pacing with an associated improvement in cardiac function. hemodynamics compared with RAA pacing through differences in the relationship between contraction of the left atrium and left ventricle during each pacing mode. Our Conclusion study demonstrated hemodynamic benefits of BiA pacing BiA pacing improves the hemodynamic status of pa- compared with RAA pacing and we consider that it is tients through prolongation of the intra-atrial conduction largely caused by changes in the relationship between the delay. The left atrial filling wave is interrupted by ventricu- contraction of the 2 chambers. The timing of ventricular lar contraction during RAA pacing, but not in BiA pacing activation from atrial pacing occurs earlier during BiA and this improvement may be the basis of the beneficial pacing than with RAA pacing and as a result, S-Q, S-A and effect of BiA pacing in preventing AF. S-A end are shortened. The left side AV interval was shortened by RAA pacing in patients with intra-atrial conduc- References tion delay, whereas BiA pacing prolonged the left side AV 1. Feinberg WM, Blackshear JL, Laupacis A, Kronmal R, Hart RG. interval (Fig4). Our study indicated that P wave duration, Prevalence, age distribution, and gender of patients with atrial fibril- S-Q, S-A and S-A end were shortened during BiA pacing. lation: Analysis and implications. Arch Intern Med 1995; 155: 469– The conduction interval from the CS to the AV node may 473. be shorter than that from the RAA to the AV node, because 2. Roy D, Talajic M, Dorian P, Connolly S, Eisenberg MJ, Green M, et al for the Canadian Trial of Atrial Fibrillation Investigators: the CS pacing site is in the lower left atrium. Ventricular Amiodarone to prevent recurrence of atrial fibrillation. N Engl J Med contraction is required for complete closure of the mitral 2000; 342: 913–920. valve and although delayed contraction following left atrial 3. Kochiadakis GE, Marketou ME, Igoumenidis NE, Chrysostomakis contraction may provoke diastolic mitral regurgitation,21 we SI, Mavrakis HE, Kaleboubas MD, et al. Amiodarone, sotalol, or did not observe it in our study. We found that 3 of the 6 pa- propafenone in atrial fibrillation: Which is preferred to maintain normal sinus rhythm? Pacing Clin Electrophysiol 2000; 23: 1883– tients in whom atrial pacing was performed had no interval 1887. of A end-I and the atrial filling wave was interrupted during 4. Komatsu T, Nakamura S, Suzuki O, Horiuchi D, Yomogida K, RAA pacing, which did not occur during BiA pacing in Okumura K. Long-term prognosis of patients with paroxysmal atrial those patients. As shown in Fig 3, obviously different fibrillation depends on their response to antiarrhythmic therapy. Circ J 2004; 68: 729–733. Doppler waveforms occurred with the 2 types of pacing. It 5. Kato T, Yamashita T, Sagara K, Iinuma H, Fu LT. Progressive is thought that the left atrial filling wave is interrupted by nature of paroxysmal atrial fibrillation: Observations from a 14-year the left ventricular contraction in patients with normal AV follow-up study. Circ J 2004; 68: 568–572. conduction and prolonged intra-atrial conduction delay 6. Fragakis N, Shakespeare CF, Lloyd G, Simon R, Bostock J, Holt D, et al. Reversion and maintenance of sinus rhythm in patients with during RAA pacing, so we guessed that cardiac function permanent atrial fibrillation by internal cardioversion followed by would be markedly improved by BiA pacing compared biatrial pacing. Pacing Clin Electrophysiol 2002; 25: 278–286. with RAA pacing. Levy et al reported that BiA pacing 7. Enjoji Y, Sugi K, Noro M, Nakao T, Tezuka N, Takami M, et al. made both left atrial and ventricular activation earlier, with Evaluation of bi-atrial pacing and single site pacing for prevention of 12 atrial fibrillation. Circ J 2002; 66: 70–74. no change in the left AV timing sequence, but we demon- 8. Gilligan DM, Fuller IA, Clemo HF, Shepard RK, Dan D, Wood MA, strated that the left side AV interval was shortened by RAA et al. The acute effects of biatrial pacing on atrial depolarization and pacing. Dobrowska-Kugacka et al13 also reported that the repolarization. Pacing Clin Electrophysiol 2000; 23: 1113–1120. atrial hemodynamics during BiA pacing were not superior 9. Fan K, Lee KL, Chiu CS, Lee JW, He GW, Cheung D, et al. Effects to those during RAA pacing and they also reported that of biatrial pacing in prevention of postoperative atrial fibrillation after coronary artery bypass surgery. Circulation 2000; 102: 755– pulmonary venous return could be reduced or left atrial 760. pressure could be increased by BiA pacing. Both those 10. Levy T, Fotopoulos G, Walker S, Rex S, Octave M, Paul V, et al. studies12,13 were in patients with left ventricular dysfunc- Randomized controlled study investigating the effect of biatrial tion in whom the left side AV conduction interval may be pacing in prevention of atrial fibrillation after coronary artery bypass grafting. Circulation 2000; 102: 1382–1387. prolonged because of inter-ventricular conduction delay, 11. Endoh Y, Nakamura A, Suzuki T, Mizuno M, Takara A, Ota Y, et al. thus masking the deficiency of RAA pacing. We excluded Clinical significance of prolonged P wave width after right atrial ap- patients with left ventricular dysfunction. Doi et al reported pendage pacing in sick sinus syndrome. Circ J 2003; 67: 485–489.

Circulation Journal Vol.69, July 2005 836 MATSUMOTO K et al.

12. Levy T, Jenkins GH, Walker S, Grieve L, Webb C, Buckley MG, et physiol 1987; 10: 409. al. Does the mechanism of action of biatrial pacing for atrial fibrilla- 17. Hagens VE, Ranchor AV, Van Sonderen E, Bosker HA, Kamp O, tion involve change in cardiac haemodynamics? Assessment by Tijssen JG, et al. Effect of rate or rhythm control on quality of life in Doppler echocardiography and natriuretic peptide measurements. persistent atrial fibrillation: Results from the Rate Control Versus Europace 2000; 2: 127–135. Electrical Cardioversion (RACE) Study. J Am Coll Cardiol 2004; 13. Dobrowska-Kugacka A, Lewicka-Nowak E, Kutarski A, Zagozdzon 43: 241–247. D, Swiatecka G. Hemodynamic effects of alternative atrial pacing 18. Flegel KM, Shipley MJ, Rose G. Risk of stroke in non-rheumatic sites in patients with paroxysmal atrial fibrillation. Pacing Clin atrial fibrillation. Lancet 1987; 1: 526–527. Electrophysiol 2003; 26: 278–283. 19. Benjamin EJ, Wolf PA, D’Agostino RB, Silbershats H, Kannel WB, 14. Doi A, Takagi M, Toda I, Yoshiyama M, Takeuchi K, Yoshikawa J. Levy D. Impact on atrial fibrillation on the risk of death: The Acute haemodynamic benefits of biatrial atrioventricular sequential Framingham heart study. Circulation 1998; 98: 946–952. pacing: Comparison with single atrial atrioventricular sequential 20. Shimizu A, Fukatani M, Tanigawa M, Mori M, Hashiba K. Intra- pacing. Heart 2004; 90: 411–418. atrial conduction delay and fragmented atrial activity in patients with 15. Ishikawa T, Sumita S, Kimura K, Kikuchi M, Kosuge M, Kuji N, et paroxysmal atrial fibrillation. Jpn Circ J 1989; 20: 1023–1030. al. Prediction of optimal atrioventricular delay in patients with im- 21. Ishikawa T, Kimura K, Nihei T, Usui T, Kashiwagi M, Ishii M. Re- planted DDD pacemakers. Pacing Clin Electrophysiol 1999; 22: lationship between diastolic mitral regurgitation and PQ intervals or 1365–1371. cardiac function in patients implanted with DDD pacemakers. Pacing 16. Rosado L, Tyers F, Cooper J. Actuarial analysis of patient survival as Clin Electrophysiol 1991; 14: 1797–1802. a function of pacemaker type pacing (abstract). Pacing Clin Electro-

Circulation Journal Vol.69, July 2005