Cardiac Resynchronisation Therapy: Is It Antiarrhythmic or Proarrhythmic?

(1)

or Proarrhythmic?

G. T

URITTO

, N. E

L

-S

HERIF

Introduction

Cardiac resynchronisation therapy (CRT) was defined in an American Heart Association Science Advisory as ‘a relatively new therapy for patients with symptomatic heart failure resulting from systolic dysfunction’ [1]. In addi- tion to improving a variety of indices of functional status, CRT may decrease morbidity and mortality. A recent meta-analysis concluded that CRT reduces mortality from progressive heart failure and suggested a trend toward longer survival in patients treated with this approach [2]. Total mortality was also reduced in two recent trials, one of which included CRT in combination with defibrillation therapy [3, 4].

Although a delayed or halted progression of cardiac dysfunction may be sufficient to prevent malignant ventricular tachyarrhythmias, there is still lingering uncertainty regarding the presence and magnitude of antiarrhyth- mic effects of CRT per se. Furthermore, there is experimental and anecdotal clinical evidence that left ventricular pacing may have proarrhythmic poten- tial, thus mandating the presence of a back-up defibrillation function in CRT devices. This review aims at reviewing evidence in favour or against an antiarrhythmic effect of CRT.

Evidence for a Proarrhythmic Effect of Left Ventricular Pacing

Basic Studies

Medina-Ravell et al. were the first to point out that the common design for

Cardiac Electrophysiology, New York Methodist Hospital and State University of New York – Downstate Medical Center, Brooklyn, NY, USA

(2)

CRT, i.e. simultaneous pacing of the right ventricular endocardium and left ventricular epicardium, is associated with a nonphysiological ventricular activation sequence. This may augment transmural heterogeneity of repolar- isation intrinsic to ventricular myocardium and, as a consequence, prolong the QT and JT intervals on ECG [5]. Their study tested this hypothesis by assessing the effects of biventricular pacing (BiVP), left ventricular epicar- dial pacing (LVEpiP), and right ventricular endocardial pacing (RVEndoP) on ventricular repolarisation, specifically on QT and JT interval and trans- mural dispersion of repolarisation (TDR), and their roles in arrhythmogene- sis in patients who received CRT. The cellular mechanisms underlying the pacing-site-dependent alterations in ventricular repolarisation were studied in an experimental model consisting of an arterially perfused rabbit LV wedge preparation in which transmembrane action potentials from endo- cardium and epicardium could be simultaneously recorded together with a transmural ECG. In the experimental preparation, TDR was defined as the difference between the longest and shortest repolarisation times across the LV wall. The authors showed that switching from endocardial to epicardial pacing resulted in a change of activation sequence between epicardium and endocardium, which was associated with an increase in QT interval and TDR without a parallel increase in endocardial and epicardial transmembrane action potential duration (APD). An increase in TDR manifested as a more positive and broader T wave on the transmural ECG. In six preparations, switching from endocardial to epicardial pacing produced a net increase in QT and TDR by 17 ± 5 and 22 ± 5 ms, respectively, at a basic cycle length of 1000 ms (P < 0.05). A more recent study examined the cellular basis for QT prolongation after reversal of the direction of activation of the LV wall [6]. Based on previous investigations documenting the contribution of M cells to TDR, this study postulated that delayed activation and repolarisation of M cells, coupled with earlier activation and repolarisation of epicardial cells, may result in QT prolongation, development of transmural heterogene- ity, and torsade de pointes after a shift from endocardial to epicardial activa- tion of the LV wall in the absence or presence of rapidly activating delayed rectifier potassium current (I

Kr

) blockade. This hypothesis was tested in a 1- dimensional mathematical model of transmural conduction and in the coro- nary-perfused canine LV wedge preparation. The results of the mathematical simulation and the experimental data confirmed that intrinsic electrical het- erogeneity exists within the ventricular myocardium and is amplified when the normal direction of activation of the ventricular wall is reversed.

Epicardial activation augments TDR because the epicardial action potential

activates and repolarises earlier while the M cells with the longest APD,

located in the deep subendocardium, activate and repolarise later compared

(3)

with endocardial activation of the ventricular wall. The additional conduc- tion delay encountered between epicardial and M regions during epicardial stimulation contributes to the amplification of TDR. M cells play a crucial role in QT prolongation, amplification of TDR, and induction of torsade de pointes that develop after a shift from endocardial to epicardial activation of the LV myocardium. The delayed activation and repolarisation of M cells, when coupled with earlier activation of repolarisation of epicardial cells, cre- ate the substrate for the development of reentry.

In another study, the roles of voltage output, interventricular delay, and pacing sites in the development of ventricular arrhythmias were investigated during BiVP or LV pacing [7]. Voltage-sensitive dye was used in eight ischaemic Langerdorff-perfused guinea-pig hearts to measure ventricular activation times and to examine conduction patterns during multisite pacing from three RV and four LV sites. Isochronal maps of RV and LV activation were plotted. Ischaemia was produced by gradually halving the perfusion output over 5 min. Pacing the RV apex and the base of the LV anterior wall was associated with the most homogeneous and rapid activation pattern (28

± 9 ms vs 41 ± 12 ms with the other configurations, P < 0.01), and no inducible arrhythmia. In six hearts, ventricular tachycardia could be induced when pacing from the right and left free walls with 20 ms of interventricular delay, at six-fold the pacing threshold output. In four hearts, simultaneous RV and LV pacing at high-voltage output induced ventricular fibrillation with complex three-dimensional propagation patterns, independently of the pacing sites. During BiVP with ischaemia, pacing at high-voltage output with a long interventricular delay is likely to induce ventricular arrhythmias, par- ticularly when left and right pacing results in a conduction pattern orthogo- nal to the orientation of the ventricular myocardial fibres.

Clinical Studies

In a study by Medina-Ravell et al., the QT interval, JT interval, and TDR were

measured in 29 patients with heart failure during RVEndoP, BiVP, and

LVEpiP [5]. The data were collected perioperatively (n = 29), 24 h after the

operation (n = 19), and during the first follow-up period, ranging from 1 to

2 weeks after the procedure (n = 12). LVEpiP and BiVP led to significant

QT and JT prolongation. LVEpiP also enhanced TDR, defined as the interval

between the peak and the end of the T wave (T

peak-end

). Frequent R-on-T

extrasystoles generated by BiVP and LVEpiP but completely inhibited by

RVEndoP occurred in four patients, of whom one developed multiple

episodes of nonsustained polymorphic ventricular tachycardia and another

suffered incessant torsade de pointes. These data suggest that, in a subpopu-

(4)

lation of patients with prolonged QT intervals, secondary to heart failure, electrolyte abnormalities, or exposure to agents with class III antiarrhythmic actions, a BiVP- or LVEpiP-dependent increase in QT interval and TDR may be a potential risk for the development of torsade de pointes. Similar occur- rences of ventricular tachyarrhythmias developing or worsening immediate- ly after CRT were the subject of two case reports [8, 9].

Evidence for an Antiarrhythmic Effect of Left Ventricular Pacing

Basic Studies

It is possible that changes described in experimental models of LV epicardial pacing may apply only to the first few hours of LV pacing in human subjects.

Previous studies have shown that changes in the transmural activation sequence can lead to enduring changes in cardiac repolarisation (i.e. ‘remod- eling’) [10, 11]. Further studies are needed to establish whether potentially proarrhythmic abnormalities that may be associated with the initiation of CRT are mitigated during long-term follow-up. At the present time, no data are available on the presence and degree of electrical remodeling after initia- tion of CRT.

Clinical Studies

There is evidence that beneficial structural and contractile LV remodeling

after CRT has a favourable effect on the frequency of spontaneous and

inducible ventricular tachyarrhythmias. The Ventak CHF Investigators

reviewed the frequency of device therapy in patients, serving as their own

controls, who were enrolled in this CRT study [12]. Of 54 patients enrolled in

the Ventak CHF trial, 32 could be analysed. Each of them completed three

blinded months programmed to BiV pacing and a second randomly assigned

3-month period of no pacing. Of the 32 patients, 13 (41%) received appropri-

ate therapy for a ventricular tachyarrhythmia at least once in the 6-month

monitoring period following implantation. Five patients (16%) had at least

one tachyarrhythmic episode while programmed to BiV pacing, whereas 11

(34%) had at least one episode while programmed to no pacing. Three

patients (9%) received therapy in both pacing periods, two with BiV pacing

only. The decrease in necessary therapy for tachycardia during the BiV pac-

ing period was statistically significant (P = 0.035). The authors concluded

that, although CRT does not obviate the need for an ICD, it might diminish

the need for appropriate tachyarrhythmia therapy. Other studies have shown

that BiVP is associated with a decrease in the inducibility of ventricular

(5)

tachycardia [13, 14]. Finally, there is anecdotal evidence that electrical storms may also be suppressed by CRT [15, 16].

Several markers of ventricular electrical vulnerability may be ameliorat- ed by CRT, which may provide indirect evidence for the antiarrhythmic effect of this therapy. Heart rate variability (HRV) was evaluated during peri- ods of pacing and no pacing in recipients of CRT [17, 18]. In a study by Adamson et al., HRV was examined in 50 patients implanted with a BiVP system who were randomised to therapy-on (n = 25) or therapy-off (n = 25) [17]. HRV was computed as the standard deviation of the atrial cycle length sensed from the system over 2 months of continuous monitoring. A comparison of HRV between CRT-on and CRT-off groups showed that HRV was higher in CRT-on than in CRT-off patients (148 ± 47 ms for CRT-on vs 118 ± 45 ms for CRT-off; P = 0.02), despite the lack of difference in mean atrial cycle length (844 ± 129 ms for CRT-on vs 851 ± 110 ms for CRT-off;

P = 0.82). These authors concluded that CRT shifts cardiac autonomic bal-

ance toward a more favourable profile that is less dependent on sympathetic activation. An increase in HRV during CRT was also reported by groups who used conventional Holter recordings to analyse HRV parameters [18].

Microvolt T-wave alternans (TWA) has been proposed as a strong inde- pendent predictor of malignant ventricular tachyarrhythmias and sudden cardiac death. We recently investigated the prevalence of TWA during dif- ferent pacing modalities in a group of patients who received CRT [19].

TWA was recorded with commercially available equipment in 16 such patients during atrial pacing (AAI) at a rate of 110/min, as well as during DDD-RV pacing and DDD-BiV pacing at the same rate and with short atri- oventricular delay, in order to obtain ventricular capture. Criteria for posi- tive TWA were: alternans > 1 min with Valt (square root of alternans power) > 1.9 μv and alternans ratio (ratio of alternans to standard devia- tion of background noise) > 3 in ≥ 1 orthogonal lead or ≥ 2 precordial leads. In this study, AAI and RV pacing resulted in a high prevalence of tachycardia-induced TWA, while BiV pacing was associated with ameliora- tion of all TWA indices (Figs. 1, 2).

Conclusions

Available evidence supports the hypothesis that CRT results in favourable

structural and electrical remodeling. Whether this effect would obviate the

need for back-up defibrillation capability in CRT devices is unclear and

should be the focus of further studies.

(6)

Fig. 1.T-wave alternans parameters obtained in 16 patients with CRT devices during atri- al pacing (AAI), right ventricular pacing (RV), and biventricular pacing (BiV) at a rate of 110/min. BiV pacing resulted in a significant improvement in T-wave alternans parameters compared to AAI or RV pacing

Fig. 2.AT-wave alternans (TWA) during RV pacing; B: TWA during BiV pacing.

Abnormal TWA values were ameliorated when switching from RV to BiV pacing

(7)

References

1. Strickberger SA, Conti J, Daoud EG et al (2005) Patient selection for cardiac resynchronization therapy. Circulation 111:2146–2150

2. Bradley DJ, Bradley EA, Baughman KL et al (2003) Cardiac resynchronization and death from progressive heart failure. A meta-analysis of randomized controlled trials. JAMA 289:730–740

3. Bristow MR, Saxon LA, Boehmer J et al (2004) Cardiac-resynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med 350:2140–2150

4. Cleland JGF, Daubert J-C, Erdmann E et al (2005) The effect of cardiac resynchronization therapy on morbidity and mortality in heart failure. N Engl J Med 352:1539–1549

5. Medina-Ravell VA, Lankipalli RS, Yan GX et al (2003) Effect of epicardial or biventricular pacing to prolong QT interval and increase transmural dispersion of repolarization. Does resynchronization therapy pose a risk for patients predisposed to long QT or torsade de pointes? Circulation 107:740–746

6. Fish JM, Di Diego JM, Nesterenko V et al (2004) Epicardial activation of the left ventricular wall prolongs QT interval and transmural dispersion of repolarization.

Implications for biventricular pacing. Circulation 109:2136–2142

7. Garrigue S, Reuter S, Efimov IR et al (2003) Optical mapping technique applied to biventricular pacing: potential mechanisms of ventricular arrhythmias occurrence.

Pacing Clin Electrophysiol 26(1 Pt 2):197–205

8. Guerra JM, Wu J, Miller JM et al (2003) Increase in ventricular tachycardia frequency after biventricular implantable cardioverter defibrillator upgrade. J Cardiovasc Electrophysiol 14:1245–1247

9. Cori AD, Bongiorni MG, Arena G et al (2005) New-onset ventricular tachycardia after cardiac resynchronization therapy. J Interv Card Electrophysiol 12:231–135 10. Geller JC, Rosen MR (1993) Persistent T-wave changes after alteration of the ventri-

cular activation sequence: new insights into cellular mechanisms of cardiac memory. Circulation 88:1811–1819

11. Libbus I, Rosenbaum DS (2003) Transmural action potential changes underlying ventricular electrical remodeling. J Cardiovasc Electrophysiol 14: 394–402

12. Higgins SL, Yong P, Sheck D et al (2000) Biventricular pacing diminishes the need for implantable cardioverter defibrillator therapy: Ventak CHF Investigators. J Am Coll Cardiol 36: 824–827

13. Zagrodzky JD, Ramaswamy K, Page RL et al (2001) Biventricular pacing decreases the inducibility of ventricular tachycardia in patients with ischemic cardiomyopathy. Am J Cardiol 87:1208–1210

14. Kowal RC, Wasmund SL, Smith ML et al (2004) Biventricular pacing reduces the induction of monomorphic ventricular tachycardia: a potential mechanism for arrhythmia suppression. Heart Rhythm 1:295–300

15. Garrigue S, Barold SS, Hocini M et al (2000) Treatment of drug refractory ventricular tachycardia by biventricular pacing. Pacing Clin Electrophysiol 23:1700–1702 16. Tanabe Y, Chinushi M, Washizuka T et al (2003) Suppression of electrical storm by

biventricular pacing in a patient with idiopathic dilated cardiomyopathy and ventricular tachycardia. Pacing Clin Electrophysiol 26(Pt 1):101–102

17. Adamson PB, Kleckner KJ, VanHout WL et al (2003) Cardiac resynchronization therapy improves heart rate variability in patients with symptomatic heart failure.

Circulation 108:266–269

(8)

18. Livanis EG, Flevari P, Theodorakis GN et al (2003) Effect of biventricular pacing on heart rate variability in patients with chronic heart failure. Eur J Heart Fail 5:175–178

19. Turitto G, Houy S, Gupta R et al (2004) Right ventricular but not biventricular pacing increases markers of ventricular electrical instability. J Am Coll Cardiol 43 (Suppl A):149A