OUP user menu

Antiarrhythmic effect of cardiac resynchronization therapy with triple-site biventricular stimulation

Michio Ogano, Yu-ki Iwasaki, Jun Tanabe, Hisato Takagi, Takuya Umemoto, Meiso Hayashi, Yasushi Miyauchi, Kyoichi Mizuno
DOI: http://dx.doi.org/10.1093/europace/eut134 1491-1498 First published online: 21 May 2013


Aims The antiarrhythmic effect of triple-site biventricular stimulation (Tri-V) is poorly understood. This study aims to evaluate the effect of cardiac resynchronization therapy (CRT) on ventricular arrhythmia (VA) with Tri-V using a single right ventricular (RV) and double left ventricular (LV) lead.

Methods and results Over a period of 3.5 years, 58 consecutive patients with New York Heart Association class II–IV heart failure, an LV ejection fraction of ≤0.35, and a QRS interval of ≥120 ms were enrolled. Acute haemodynamic responses to dual-site biventricular stimulation (Bi-V) and Tri-V were evaluated by assigning patients to a Bi-V or Tri-V group. Electrocardiogram parameters [QT interval, JT interval, and transmural dispersion of repolarization (TDR)] were measured over time after CRT. Spontaneous VA detected by telemetry was reviewed and confirmed. During a mean follow-up of 481 days after implantation, VA occurred in 2 of 22 patients in the Tri-V group and 14 of 36 patients in Bi-V group. Triple-site biventricular stimulation was thus associated with a decreased VA risk (P = 0.044). Multivariate Cox analysis showed that Tri-V pacing prevented arrhythmic events as compared with Bi-V pacing (hazard ratio, 0.13; 95% confidence interval, 0.029–0.610; P = 0.009). Ventricular repolarization indices at 6 months were significantly shortened in Tri-V compared with Bi-V (QTc, −23.6 vs. −14.1%, P = 0.008; JTc, −21.4 vs. −7.7%, P = 0.005; TDRc, −39.9 vs. −17.0%, P < 0.001).

Conclusion Compared with Bi-V, Tri-V reduced VA during long-term follow-up. Improvements in repolarization parameters may result in antiarrhythmic effects.

  • Cardiac resynchronization therapy
  • Multi-site pacing
  • Ventricular arrhythmia
  • Antiarrhythmia
  • Ventricular repolarization.

What's new?

  • Cardiac resynchronization therapy (CRT) with triple-site biventricular stimulation (Tri-V), involving double left ventricular (LV) and single right ventricular pacing, showed antiarrhythmic effect compared with CRT with conventional dual-site biventricular stimulation (Bi-V), in spite of the fact that extent of LV volume reduction by CRT was similar in patients with Tri-V and Bi-V.

  • Ventricular repolarization indices: QT interval, JT interval, and transmural dispersion of repolarization, were shortened in patients with Tri-V compared with Bi-V 6 months after implantation.

  • Reduction of dispersion of refractoriness by multi-site pacing may play an important role in antiarrhythmic effect by CRT.


Cardiac resynchronization therapy (CRT) is a well-established treatment for patients with symptomatic heart failure, depressed left ventricular (LV) ejection fraction, and widened QRS.1,2 Insertion of a right ventricular (RV) lead to pace the apex of the RV and introduction of an LV lead into the coronary sinus (CS) to pace the LV epicardium are currently considered standard procedures for resynchronization therapy. The non-physiological LV epicardial activation sequence can augment transmural heterogeneity of repolarization and consequently prolong the QT and JT intervals and transmural dispersion of repolarization (TDR) on the electrocardiogram (ECG).3,4 There have been sporadic case reports of precipitation of ventricular tachycardia (VT) or ventricular fibrillation (VF) with CRT.5,6 Another limitation of CRT is that up to 30% of patients do not obtain sufficient LV reverse remodelling.7

Recent reports suggest that CRT with triple-site biventricular stimulation (Tri-V), involving double LV and single RV pacing, is safe and more beneficial than standard CRT [dual-site biventricular stimulation (Bi-V)] in terms of reverse remodelling of the LV.810 The magnitude of reduction in LV volume after CRT is related to the reduction in the risk of VT/VF.11 However, no data have been reported regarding the potential beneficial effects of reducing LV volume in VT/VF in patients with Tri-V. In the present study, we aimed to investigate the association between Tri-V and the occurrence of subsequent VT/VF at long-term follow-up.



From July 2008 to December 2011, a total of 58 consecutive patients with New York Heart Association (NYHA) functional class II–IV heart failure were enrolled. All patients had ischaemic or non-ischaemic cardiomyopathy in association with an LV ejection fraction of ≤0.35 and a QRS interval of ≥120 ms. All patients were receiving optimal medical heart failure therapy, which included a stable dose of an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker, diuretics, and a β-blocker. The present study was approved by our institution's ethics committee.

Before device implantation, the following baseline demographic data were collected: medical history—history of VT/VF, syncope or cardiac arrest; 12-lead ECG results; NYHA functional class; and echocardiographic parameters, including left ventricular end-systolic volume (LVESV), left ventricular end-diastolic volume (LVEDV), and LV ejection fraction. Clinical characteristics are presented in Table 1.

View this table:
Table 1

Baseline patient characteristics

CharacteristicsTotal (n = 58)
Male sex (%)43 (74.3)
Age (years)69 ± 13
Secondary prevention for ICD (%)14 (24.1)
Ischaemic aetiology (%)24 (41.4)
Atrial fibrillation (%)6 (10.3)
LVEF (%)25.5 ± 10.4
NYHA class (%)
 II13 (22.4)
 III31 (53.4)
 IV14 (24.1)
QRS duration (ms)150.8 ± 32.9
LVEDV (mL)236.0 ± 107.1
LVESV (mL)183.3 ± 102.1
Medication at baseline (%)
 ACE-I/ARB51 (87.9)
 β-Blockers44 (75.9)
 Diuretics43 (74.1)
  • ICD, implantable cardioverter defibrillator; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume; ACE-I, angiotensin-converting-enzyme inhibitor; ARB, angiotensin-receptor blocker.

Temporary pacing study for acute haemodynamic measurements

Written informed consent was obtained, and an invasive temporary pacing study was performed before CRT defibrillator implantation. For haemodynamic measurements, a dual-transducer pressure catheter (Pressure Tip Catheter, CD leycom) was introduced into the LV. Temporary pacing leads were placed into the high right atrium and RV apex. Coronary sinus was cannulated with an 8.5 Fr pre-shaped SL2 sheath (St Jude Medical). Two 3.5 Fr over-the-wire-type pacing catheters (InterNova Monorail Catheter, InterNova) with 0.014 inch guidewires were placed in each CS branch eligible for LV pacing.

Baseline LV-dP/dtmax was measured during pacing in the AAI mode (10b.p.m. above the intrinsic rate). With the same baseline pacing cycle length, pacing in the DDI mode [atrioventricular (AV) delay was set at 100 ms below baseline AV delay to ensure consistent ventricular capture] was performed from the RV and single LV sites (dual-site ventricular pacing) and the RV and two LV sites (triple-site ventricular pacing) simultaneously. For patients with atrial fibrillation, the pacing mode was programmed to VVI with a pacing rate of 15 b.p.m. above the average spontaneous heart rate. The LV dP/dtmax was calculated electrically from every heart beat for a period of at least 10 s during steady-state conditions.

Implantation of the cardiac resynchronization therapy defibrillator system

Enrolled patients were divided into two groups (Tri-V group or Bi-V group) based on the results of the acute haemodynamic study. If the patients who obtained better haemodynamic response (LV-dP/dtmax) in triple-site pacing rather than dual-site ventricular pacing, they were assigned to Tri-V group. On the other hand in the patient with better haemodynamic response in dual-site ventricular pacing rather than in triple-site, they were assigned to Bi-V group. During temporary pacing study the patient without suitable CS branches for triple-site pacing were assigned to Bi-V group. Atrial and defibrillation leads were implanted in a conventional manner. Left ventricular leads were implanted into the target CS branches that showed the best haemodynamic improvement during pacing study. The size and type (unipolar or bipolar) of lead implanted were determined by the anatomy of the CS branches. All leads were connected to a CRT defibrillator (Protecta XT or Consulta, Medtronic). For patients who received Tri-V, both LV leads were connected to the LV port of the CRT defibrillator using parallel Y-adaptors (Lead Adaptor 2872, Medtronic).

Device programming and follow-up

For primary prevention, a cut-off VT rate of 188 b.p.m. was programmed with a series of anti-tachycardia pacing (ATP) followed by shocks. Anti-tachycardia pacing was started using burst pacing with 15 pulses at 88% of the measured VT cycle length. Secondary therapy was shock at a level of the defibrillation threshold plus at least 10 J. The remaining therapies were maximal energy shocks. For secondary prevention, a cut-off VT rate of 20 b.p.m. lower than the clinically documented VT was selected. The cut-off VF rate was programmed at 200 b.p.m. for all patients.

Patients were regularly followed up every 3 months and surface ECG, NYHA functional class, and device checks were performed. In case of shock delivery or syncope, emergent device telemetry check was performed. Spontaneous ventricular arrhythmias (VAs) (monomorphic VT, polymorphic VT, and VF) detected by implantable cardioverter defibrillator (ICD) telemetry were interpreted by two experienced physicians. The impact of CRT on echocardiography outcome was evaluated using follow-up data at 6 months.

Index of ventricular repolarization

The QT interval, JT interval, and TDR in lead II were measured before CRT implantation and 1 day, 3 months, and 6 months after CRT implantation. Patients with atrial fibrillation were excluded from these measurements. The QT interval was defined as the time interval between the initial deflection of the QRS and the point at which a tangent drawn to the steepest portion of the terminal part of the T-wave crossed the isoelectric line. Similarly, the JT interval that excludes the QRS component was measured as the time from J-point to the end of the T-wave. Transmural dispersion of repolarization was defined as the interval between the peak to the end of the T-wave. Corrected QT (QTc), corrected JT (JTc), and corrected TDR (TDRc) values were calculated using Bazett's formula.12 All ECG measurements were performed by a blinded observer.

Statistical analysis

The Student's t-test or Welch test was used to compare continuous data between two groups, as appropriate. Categorical variables were expressed as absolute counts and percentages; these variables were compared using the χ2 test or Fisher's exact test, as appropriate. Comparisons of continuous data at baseline and follow-up points were performed using a paired Student's t-test for paired data and one-way analysis of variance (ANOVA) coupled with post hoc analysis using Bonferroni's test between three or more groups. Statistical analysis of the time course of the QT and JT intervals and TDR rate changes between Tri-V and Bi-V patients was performed using two-way ANOVA coupled with post hoc analysis using Bonferroni's test.

Cumulative appropriate ICD event rates from implantation until the last follow-up were calculated using the Kaplan–Meier method. The log-rank tests for time to event data were used for statistical comparison the between two groups. Univariate Cox regression analysis was used for the analysis of predictors of VA events. All variables in the univariate model that influenced the probability of an arrhythmic event with a significance level of P ≤ 0.10 were subsequently entered into multivariate Cox regression analysis. The results were expressed as hazard ratio (HR)/unit change ±95% confidence interval (CI). All tests were double-sided. P values of <0.05 were considered statistically significant. Data were expressed as mean ± standard deviation. All statistical analyses were performed using SPSS for Windows, version 19.0 (IBM SPSS Statistics, IBM Corp.).


Haemodynamic response and device implantation

All 58 patients underwent temporary pacing study before device implantation. The LV-dP/dtmax was successfully measured in all 58 patients using dual-site ventricular pacing and in 40 patients (69.0%) with triple-site ventricular pacing. Twenty-two patients (37.9%) obtained a better haemodynamic response with triple-site ventricular pacing than with dual-site ventricular pacing, and assigned to Tri-V group. Triple-site biventricular stimulation was successfully implanted in all of these patients (Figure 1). The other 18 patients had better haemodynamic response in Bi-V than that in Tri-V. Eighteen patients who did not have accessible suitable CS branches for tri-ventricular pacing (including high LV output threshold, or phrenic nerve stimulation) were assigned to Bi-V group. In total 36 patients were assigned to Bi-V group. Except for the QRS duration, there were no differences in the baseline patient characteristics of the Bi-V group between accomplished Tri-V and unaccomplished Tri-V pacing during the temporary pacing study (Table 2).

View this table:
Table 2

Baseline patient characteristics separated according to accomplishment of Tri-V temporary pacing study

CharacteristicsAccomplished patient (N = 18)Unaccomplished patient (N = 18)P value
Male sex (%)15 (83.3)11 (61.1)0.137
Age (years)68 ± 1473 ± 130.762
Secondary prevention (%)4 (22.2)3 (16.7)0.674
Ischaemic aetiology (%)8 (44.4)7 (38.9)0.735
LVEF (%)28.3 ± 10.426.6 ± 10.20.753
NYHA class (n, %)0.540
 II5 (27.8)6 (33.3)
 III11 (61.1)8 (44.4)
 IV2 (11.1)4 (22.2)
QRS duration (ms)137.6 ± 21.2161.3 ± 38.90.036
EDV (mL)212.4 ± 103.3219.3 ± 111.10.721
ESV (mL)158.4 ± 93.4167.2 ± 104.90.885
Baseline dP/dt1090.8 ± 283.41013.6 ± 355.40.734
  • LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; EDV, end-diastolic volume; ESV, end-systolic volume.

Figure 1

Distribution of LV lead location. The distribution of the LV pacing lead position in triple-site pacing group (A) and dual-site pacing group (B). Tri-V, triple-site biventricular stimulation; Bi-V, dual-site biventricular stimulation; A, anterior site; AL, anterolateral site; L, lateral site; PL, posterolateral site; P, posterior site.

The average procedure time for temporary pacing study was 69.1 ± 20.5 min. Intraoperative complications were not observed in any of the patients. The procedure time for the implantation was slightly longer in the Tri-V group than in the Bi-V group, but the difference was not statistically significant (181.4 ± 45.5 vs. 164.4 ± 31.1min, P = 0.21). In two patients with Bi-V pacing, phrenic nerve stimulation due to lead dislodgement occurred within 2 days after implantation; re-positioning of the leads successfully abolished the symptoms.

Cardiac resynchronization therapy response and ventricular arrhythmic events

The total mean follow-up duration was 481 ± 335 days and was comparable in the two groups (Tri-V group, 477 ± 327 days; Bi-V group, 483 ± 344 days; P = 0.945). Five deaths occurred in the overall population (three in the Tri-V group vs. two in the Bi-V group, P = 0.357). One patient with Tri-V died due to end-stage heart failure. The other two patients with Tri-V and two patients with Bi-V died from non-cardiac causes.

Overall, both clinical symptoms and echocardiographic parameters improved at 6 months after CRT. The mean NYHA functional class improved from 3.0 ± 0.7 to 1.9 ± 0.7 (P < 0.001). Furthermore, a decrease in LVEDV from 236.0 ± 107.1 to 203.0 ± 95.9 mL (P < 0.001), a decrease in LVESV from 183.5 ± 102.1 to 150.7 ± 90.3 mL (P < 0.001), and a consequential increase in LV ejection fraction from 25.5 ± 10.4 to 29.7 ± 12.5% (P = 0.007) were noted.

Appropriate ICD therapies, including shock and ATP were implemented for 16 patients (27.6%) over the follow-up period. No inappropriate ICD therapies were noted for any patients. Polymorphic VT or VF was documented in 1 patient with Tri-V and 3 patients with Bi-V (4.5 vs. 8.3%, P = 1.000), and monomorphic VT was documented in 1 patient with Tri-V and 11 patients with Bi-V (4.5 vs. 30.6%, P = 0.021). Figure 2 shows the Kaplan–Meier curve for appropriate ICD therapy (log-rank: mean 961 days, 95% CI: 790–1103, P = 0.044). In univariate Cox regression analysis, patients who received appropriate ICD therapies were more likely to have received ICD implantation for secondary prevention (HR = 4.396; 95% CI, 1.641–11.781; P = 0.003) and less likely to have received CRT with Tri-V (HR = 0.245; 95% CI, 0.055–1.080; P = 0.063) (Table 3). In the multivariate Cox regression analysis, independent predictors of VA events were ICD implantation for secondary prevention (HR = 7.499; 95% CI 2.664–21.108; P < 0.001) and CRT with Tri-V (HR = 0.133; 95% CI 0.029–0.610; P = 0.009) (Table 3). Cardiac resynchronization therapy with Tri-V pacing was an independent predictor of prevention of VA events.

View this table:
Table 3

Cox proportional HR model to predict VA event

HR95% CIP valueHR95% CIP value
Secondary prevention for ICD4.3961.641–11.7810.0037.4992.664–21.108<0.001
Ischaemic aetiology2.1120.790–5.7030.138
CRT with Tri-V0.2450.055–1.0800.0630.1330.029–0.6100.009
NYHA class1.0880.523–2.2640.821
QRS duration0.9870.969–1.0050.146
Baseline dP/dt1.0010.999–1.0030.217
  • HR, hazard ratio; CI, confidence interval; CRT, cardiac resynchronization therapy; Tri-V, triple-site biventricular stimulation; ICD, implantable cardioverter defibrillator; NYHA, New York Heart Association; LVEF, left ventricular ejection fraction; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume.

Figure 2

Kaplan–Meier curve for appropriate ICD therapy. Kaplan–Meier curve for appropriate ICD events in patients undergoing CRT with Tri-V (solid line) and Bi-V (broken line). Tri-V, triple-site biventricular stimulation; Bi-V, dual-site biventricular stimulation.

Differences between patients with triple-site and dual-site ventricular pacing

We compared baseline characteristics between the Tri-V and Bi-V groups, and revealed that the Tri-V group had a tendency towards a lower LV ejection fraction, increased LV volume, lower baseline dP/dt, longer QTc, JTc, and TDRc, and higher frequency of patients with NYHA functional class IV and secondary prevention for ICD (Table 4). There were no differences in the distribution of the LV lead position between first LV lead in the Tri-V and Bi-V groups (left anterior oblique view; P = 0.141, right anterior oblique view; P = 0.845). All second LV leads in the Tri-V group were implanted at the anterior or anterolateral site (Figure 1).

View this table:
Table 4

Baseline patient characteristics separated according to CRT with Tri-V and Bi-V

CharacteristicsTri-V (N = 22)Bi-V (N = 36)P value
Male sex (%)17 (77.3)26 (72.2)0.670
Age (years)68 ± 1271 ± 130.382
Secondary prevention (%)8 (36.4)7 (19.4)0.153
Ischaemic aetiology (%)9 (40.9)15 (41.7)0.955
Atrial fibrillation (%)4 (18.2)3 (8.3)0.409
LVEF (%)22.3 ± 10.027.4 ± 10.20.064
NYHA class (n, %)
QRS duration (ms)153.5 ± 33.6149.1 ± 32.90.628
EDV (mL)268.9 ± 103.2215.9 ± 105.80.067
ESV (mL)216.8 ± 102.0162.8 ± 98.00.050
Baseline dP/dt905.1 ± 250.41055.7 ± 315.40.070
Tri-V (N = 18)Bi-V (N = 33)P value
QTc (ms)477.3 ± 54.9444.4 ± 52.90.044
JTc (ms)291.1 ± 40.7282.6 ± 38.90.545
TDRc (ms)120.8 ± 24.8102.8 ± 15.70.002
  • Bi-V, dual-site ventricular stimulation; TDR, transmural dispersion of repolarization; Tri-V, triple-site biventricular stimulation; NYHA, New York Heart Association; LVEF, left ventricular ejection fraction; EDV, end-diastolic volume; ESV, end-systolic volume.

At the 6-month follow-up, the Tri-V and Bi-V groups showed significant improvement in NYHA classification (Tri-V: 3.3 ± 0.7 vs. 2.1 ± 1.0, P < 0.01; Bi-V: 2.9 ± 0.7 vs. 1.8 ± 0.6, P < 0.01), and reduction in LVEDV (Tri-V: 264.7 ± 96.5 vs. 230.0 ± 100.2 mL, P = 0.015; Bi-V: 215.9 ± 105.8 vs. 188.4 ± 89.1 mL, P = 0.001), and LVESV (Tri-V: 213.6 ± 98.3 vs. 179.7 ± 98.3 mL, P = 0.024; Bi-V: 162.8 ± 98.0 vs. 135.7 ± 80.5 mL, P = 0.002) as compared with the baseline (Table 5). There were no significant differences at 6 months between the Tri-V and Bi-V groups with respect to these parameters (NYHA classification: −1.2 ± 0.8 vs. −1.1 ± 0.7, P = 0.725, LVEDV: −34.8 ± 56.8 vs. −27.4 ± 46.5 mL, P = 0.608, LVESV: −33.9 ± 59.8 vs. −27.1 ± 47.8 mL, P = 0.651). We confirmed that after CRT, patients without VA events showed significant QRS shortening (154.5 ± 35.4 vs. 139.2 ± 19.2 ms, P = 0.007) and LV volume reduction (LVEDV; 237.0 ± 115.0 vs. 200.6 ± 103.5 mL, P < 0.001, LVESV; 185.1 ± 110.4 vs. 149.3 ± 99.0 mL, P < 0.001) with increasing LVEF (24.8 ± 10.2 vs. 30.8 ± 14.1%, P = 0.003). In contrast, CRT did not show significant improvement in electrocardiographic and echocardiographic parameters in patients with VA events (Table 6).

View this table:
Table 5

Clinical and echocardiographic parameters at 6-month follow-up in CRT Tri-V and Bi-V

CharacteristicsTri-V (N = 19)Bi-V (N = 36)
Baseline6 monthsP valueBaseline6 monthsP value
NYHA class (%)
 I05 (26.3)010 (27.8)
 II2 (10.5)9 (47.4)11 (30.6)24 (66.7)
 III10 (52.6)3 (15.8)19 (52.8)2 (5.6)
 IV7 (36.8)2 (10.5)6 (16.7)0
3.3 ± 0.72.1 ± 1.0<0.0012.9 ± 0.71.8 ± 0.6<0.001
LVEF (%)21.3 ± 8.824.7 ± 9.50.13727.4 ± 10.232.0 ± 13.30.037
QRS duration (ms)152.5 ± 35.2144.8 ± 20.20.306149.1 ± 32.9134.2 ± 19.40.005
LVEDV (mL)264.7 ± 96.5230.0 ± 100.20.015215.9 ± 105.8188.4 ± 89.10.001
LVESV (mL)213.6 ± 98.3179.7 ± 98.30.024162.8 ± 98.0135.7 ± 80.50.002
  • Bi-V, dual-site ventricular stimulation; Tri-V, triple-site biventricular stimulation; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume.

View this table:
Table 6

Ventricular arrhythmic events and LV reverse remodelling

CharacteristicsPatients with ventricular arrhythmic events (N = 16)Patients without ventricular arrhythmic events (N = 39)
Baseline6 monthsP valueBaseline6 monthsP value
LVEF (%)26.6 ± 9.926.3 ± 6.90.32724.8 ± 10.230.8 ± 14.10.003
QRS duration (ms)140.4 ± 26.8134.9 ± 22.80.903154.5 ± 35.4139.2 ± 19.20.007
LVEDV (mL)222.5 ± 75.1207.9 ± 69.50.055237.0 ± 115.0200.6 ± 103.5<0.001
LVESV (mL)168.9 ± 71.2154.8 ± 59.10.104185.1 ± 110.4149.3 ± 99.0<0.001
  • LVEF, left ventricular ejection fraction; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume.

Figure 3 shows the time course of the indices of ventricular repolarization in both groups. One day after implantation, the QTc and JTc intervals significantly increased compared with baseline values in the Bi-V group, whereas the Tri-V group showed no QTc or JTc interval prolongation. At 3- and 6-month follow-ups, QTc and JTc intervals and TDRc significantly reduced in both groups, except the JTc interval in the Bi-V group. Triple-site biventricular stimulation vs. Bi-V comparisons at each time point revealed a significant shortening of QTc, JTc, and TDRc at 3 months after CRT (QTc, −23.7 vs. −13.9% compared with baseline, P = 0.031; JTc, −20.9 vs. −6.9%, P = 0.010; TDRc, −36.2 vs. −12.9%, P < 0.001), all favouring Tri-V. The advantage of Tri-V over Bi-V with respect to these parameters remained at 6 months after CRT (QTc, −23.6 vs. −14.1%, P = 0.008; JTc, −21.4 vs. −7.7%, P = 0.005; TDRc −39.9 vs. −7.0%, P < 0.001).

Figure 3

Time-course change in QT and JT interval and TDR. Time-course change in QT and JT intervals and TDR at 1 day, 3 months, and 6 months after CRT implantation [*P < 0.05, **P < 0.01 with respect to the baseline state (baseline state = 100%); †P < 0.05, ††P < 0.01 vs. Bi-V group]. TDR, transmural dispersion of repolarization.


The findings of the present study can be summarized as follows: compared with conventional Bi-V, Tri-V (i) reduced VA events and (ii) resulted in a more homogeneous ventricular repolarization over time.

Antiarrhythmic effect of resynchronization therapy

The mechanism of the antiarrhythmic effect of CRT has two contributing factors. One involves LV reverse remodelling. In the MADIT-CRT study, a reduction in VA events was correlated with the degree of ventricular reverse remodelling induced by CRT.11 The other factor involves the reduction of dispersion of refractoriness. In 1988, Restivo et al.13 reported that stimulation at two ventricular sites prevented the induction of VT/VF in the post-infarction heart. Pre-excitation of the slow conduction area reduced the intraventricular conduction delay and dispersion of refractory periods, and provided more homogeneous ventricular repolarization. Kowal et al.14 reported that biventricular pacing significantly reduced the induction of monomorphic VT in 18 patients. Reductions in VT might be enhanced by the proximity of the LV lead to the re-entrant site. However, to obtain sufficient antiarrhythmic effects through the reduction of dispersion of refractoriness, the pacing site should be positioned in the slow conduction area that is responsible for re-entrant tachyarrhythmia. Inappropriate site pacing has no antiarrhythmic effect, but it might increase the occurrence of VT/VF. For patients with initiation of monomorphic VT shortly after CRT, the LV lead was placed in a site remote from the origin of VT.15 Additional pacing might directly influence the penetration of the first beat into the re-entrant circuit, leading to antiarrhythmic effects in some patients but pro-arrhythmic effects in others.

In the present study, both Tri-V and Bi-V reduced the QTc and JTc interval and TDRc at 3 and 6 months after CRT. The effect of CRT on ventricular repolarization has been debated.3,16 However, few studies have focused on the change in ventricular repolarization over time. Recently, Chen et al.17 has reported that alternation of ventricular epicardial activation potential duration during CRT is significantly correlated with the effect of CRT and may explain the incidence of arrhythmias. Wecke et al.18 reported that biventricular pacing led to temporary increases in the QTc interval after CRT; the QTc interval then decreased 2 weeks after CRT. These results are consistent with the results of the Bi-V group in the present study. In this study, we first demonstrated the time course of ventricular repolarization parameters in Tri-V. At 3 months after CRT, ventricular repolarization indices were shortened in Tri-V compared with Bi-V. These effects persisted through 6 months. It is speculated that Tri-V provided more pre-excitation of the larger ventricular myocardium, more homogeneous repolarization, and a further decrease in the dispersion of repolarization than did conventional Bi-V. Triple-site biventricular stimulation might effectively reduce the ventricular dispersion of refractoriness, leading to prevention of VA events. In contrast to the observed increase in arrhythmic events over time in the Bi-V group, the Tri-V group had fewer arrhythmic events 3 months after CRT.

Left ventricular reverse remodelling could contribute to the observed antiarrhythmic effects, although there were no significant differences associated with LV reverse remodelling between the groups at the 6-month follow-up. Advanced heart failure in Tri-V patients at baseline might offset any effects of LV reverse remodelling by Tri-V.

In this study, significant LV reverse remodelling was observed only in patients without VA events. Therefore, in addition to the reduction of dispersion of LV refractoriness, LV reverse remodelling might play an important role in the prevention of VA events. Further studies are required to elucidate the causal relationship between antiarrhythmic effect with CRT and Tri-V.

Recent basic science and clinical studies have shown that reversing activation sequences through the LV wall enhances TDR and could induce polymorphic VT or VF.3,4,19 In a small subgroup of patients, polymorphic VT or VF was observed after CRT initiation.5,6 However, large-scale trials of CRT have not demonstrated pro-arrhythmic events. Therefore, epicardial pacing for heart failure might induce a pro-arrhythmic effect, mainly because of predisposing factors such as myocardial ischaemia, electrolyte imbalance, pre-existing prolonged QT caused by antiarrhythmic agents, or because of the influence of the autonomic nervous system.19 Prominent reduction of TDR by Tri-V might contribute to the fewer arrhythmic events in Tri-V compared with Bi-V.

Triple-site ventricular pacing

Recent randomized trials have demonstrated that Tri-V is more effective for LV reverse remodelling than Bi-V.8,9 However, in the present study, the magnitude of LV reverse remodelling did not differ between Tri-V and Bi-V. One reason is that patients in the Tri-V group had more advanced heart failure and larger LV dimensions than did those in the Bi-V group.

In this study, the second LV lead for Tri-V was located mostly in the great cardiac vein (GCV). In fact, it is not suitable to place the LV lead in the GCV for Bi-V pacing.20 However, no data indicating the best second LV lead position for Tri-V pacing have been reported. Further studies are required for determining the most suitable second LV position for CRT with Tri-V pacing.


Since classification of patients into the Tri-V and Bi-V groups was based on the results of temporary pacing study, this study was non-randomized. Actually, patients were assigned into either Tri-V or Bi-V group according to the result of temporary pacing study. Triple-site biventricular stimulation group had a tendency to have advanced heart failure, although there were no statistically significant differences between the two groups. The outcome could have been biased by patient selection. However, despite the more advanced state of heart failure in patients enrolled in the Tri-V group, a greater antiarrhythmic effect was achieved using Tri-V compared with Bi-V.

In the present study, temporary pacing study and acute haemodynamic analyses were performed. Generally, the optimal LV lead position is the latest activation site, as determined by echocardiography.21 However, the area of latest activation was insufficient to predict the optimal sites in cases where the area was not in proximity to suitable veins, did not have appropriate output or threshold, or was in the region of phrenic nerve stimulation.22 Likewise, when Tri-V is used, the wall motion pattern can be complicated by pacing, preventing assessment of the optimal lead position through echocardiography. Although the temporary pacing study and acute haemodynamic studies place some burden on patients, these procedures were efficiently performed just before CRT implantation in this study without any complications. The data obtained were used to tailor LV lead number and position to suit each patient.23

We chose only lead II for the repolarization parameter measurements. Repolarization parameter measurements from a single lead might influence the accuracy of ventricular repolarization.

Finally, our study is limited by its single-centre design and small study population. Further multicentre randomized studies with large numbers of patients are warranted to confirm these findings.


Patients who received CRT with Tri-V experienced fewer VA events over long-term follow-up than did those who received CRT with Bi-V. Shortening of ventricular repolarization indices by Tri-V indicates homogeneous ventricular repolarization, which might play an important role in the antiarrhythmic effect of Tri-V.


The authors thank Kunito Shiba and Hidekazu Kawanaka for technical assistance during temporary pacing study; Koji Nakamura for assistance in device operation; and Michitaka Katoh for assistance in patient follow-up.

Conflict of interest: none declared.


View Abstract