Abstract
Background
To evaluate the impact of changes in the filtered QRS duration (fQRS) on signal‐averaged electrocardiograms (SAECGs) from pre‐ to postimplantation on the clinical outcomes in nonischemic heart failure (HF) patients under cardiac resynchronization therapy (CRT).
Methods
We studied 103 patients with nonischemic HF and sinus rhythm who underwent CRT implantation. SAECGs were obtained within 1 week before and 1 week after implantation and narrowing fQRS was defined as a decrease in fQRS from pre‐ to postimplantation. Echocardiography was performed before and 6 months after CRT implantation. The primary outcome was death from any cause. The secondary outcomes were hospitalization due to worsened HF and occurrence of ventricular tachyarrhythmias.
Results
Of the 103 CRT patients, 53 (51%) showed narrowing fQRS. Left ventricular end‐diastolic volume and end‐systolic volume were significantly reduced (both p < .001), and the left ventricular ejection fraction was significantly increased (p < .001) after CRT in patients with narrowing fQRS, but not in patients with nonnarrowing fQRS. During a median follow‐up period of 33 months, patients with narrowing fQRS exhibited better survival than patients with nonnarrowing fQRS (p = .007). A lower incidence of hospitalization due to worsened HF (p < .001) and a lower occurrence of ventricular tachyarrhythmias (p = .071) were obtained in patients with narrowing fQRS. After adjusting for confounding variables, narrowing fQRS was associated with a low risk of mortality (HR 0.27, p = .006).
Conclusion
Our results suggested that narrowing fQRS on SAECG after CRT implantation predicts LV reverse remodeling and long‐term outcomes in nonischemic HF patients.
Keywords: cardiac resynchronization therapy, heart failure, mortality, QRS duration, signal‐averaged electrocardiogram, ventricular arrhythmias
1. INTRODUCTION
Cardiac resynchronization therapy (CRT) can decrease mortality and morbidity in congestive heart failure (HF) patients with impaired left ventricular (LV) function and cardiac dyssynchrony (Brignole et al., 2013). HF patients with nonischemic etiology, those with left bundle branch block (LBBB) morphology and those with wider QRS duration could derive significant benefits for mortality, morbidity and LV function from this therapy (Brignole et al., 2013).
Reverse LV remodeling by CRT, which is assessed as a response by echocardiography, predicts clinical outcomes, including symptoms, exercise tolerance, and death or HF events, and reduces disease progression. Previous reports have suggested that narrowing QRS duration after CRT implantation is associated with enhanced reverse LV remodeling (responder to CRT) (Molhoek et al., 2004; Lecoq et al., 2005; Rickard et al.,2011, 2013; Karaca et al., 2016). The predictors of response to the CRT (PROSPECT)‐ECG trial suggested that the difference between the biventricular pace and preimplantation QRS duration can predict improvement of the clinical composite score (Hsing et al., 2011). However, few studies have reported the predictive role of changes in the QRS duration in the early period after implantation relative to long‐term clinical outcomes in CRT patients.
Signal‐averaged electrocardiogram (SAECG) is a surface high‐resolution electrocardiography technique that detects low‐amplitude wave forms within the terminal portion of QRS waves (ventricular late potentials) that cannot be detected by a standard 12‐lead electrocardiogram (12‐lead ECG). Filtered QRS duration (fQRS) on SAECG could predict the arrhythmic substrate, which leads to the development of ventricular tachyarrhythmias and is the most robust component related to outcomes in infarcted patients (Goldberger et al., 2008). Although controversy exists, abnormal SAECGs, most of which reflect a prolonged fQRS, could predict ventricular tachyarrhythmias, sudden cardiac death (SCD) or total mortality in patients with nonischemic‐dilated cardiomyopathy (Mancini, Wong, & Simson, 1993; Goldberger et al., 2008).
It is difficult to visually measure the specific QRS duration using a 12‐lead ECG. Recently, some reports have suggested that SAECGs may be a more useful tool for clarification of CRT candidates and their responses to CRT (Andrikopoulos et al., 2009; Mohajer et al., 2014). The aim of this study is to evaluate the impact of changes in fQRS on SAECGs from pre‐ to postimplantation on clinical outcomes in nonischemic HF patients who received CRT.
2. METHODS
All patients with nonischemic HF and sinus rhythm who underwent new implantation of a CRT device from January 2004 to December 2015 at Tokyo Women's Medical University Hospital were retrospectively included in this analysis. Nonischemic etiology was defined as the absence of coronary artery disease, which was confirmed by coronary angiography. The LV ejection fraction (LVEF) was calculated by echocardiography. The indications for CRT were New York Heart Association (NYHA) functional class II to IV despite adequate medical treatment, LVEF ≤35%, and QRS duration ≥120 ms or permanent right ventricular pacing according to the guidelines of the European Society of Cardiology or Japanese Circulation Society (Brignole et al., 2013; JCS Joint Working Group, 2013). Patients with a history of sustained ventricular tachycardia (VT) or ventricular fibrillation (VF) or those who survived cardiac arrest received CRT with an ICD (CRT‐D) implantation for secondary prevention of SCD. Patients who were also indicated for ICD as primary prevention of SCD received CRT‐D. All other patients received CRT without an ICD. We excluded patients who received intravenous inotropes or mechanical support such as intra‐aortic balloon pumping and percutaneous cardiopulmonary support.
After device implantation, the LV lead position was determined using biplane fluoroscopy. In all patients, atrioventricular (AV) and interventricular (VV) delay optimization were performed using echocardiography. SAECGs were obtained within 1 week before and 1 week after implantation in all patients. This study was approved by the institutional review board of Tokyo Women's Medical University.
2.1. Clinical outcomes
The primary outcome was death from any cause. The secondary outcomes were hospitalization due to worsened HF and occurrence of ventricular arrhythmias from the implantation of CRT to the first event. Worsened HF was defined by signs and symptoms, such as dyspnea, rales, and ankle edema, and the need for treatment with diuretics, vasodilators, positive inotropic drugs, or an intraaortic balloon pump. Ventricular tachyarrhythmias were defined as VT/VF requiring ICD therapy or external defibrillation, and intravenous antiarrhythmics such as amiodarone.
2.2. Follow‐up
Follow‐ups were performed every 3–6 months up to December 2015 at our pacemaker or ICD clinic. The median follow‐up period was 33 months, and the range was 5–118 months. Data on ventricular tachyarrhythmia occurrence requiring ICD therapy, including both shock and antitachycardia pacing, were obtained by reviewing the event details and electrograms stored on CRT disks. Only episodes of VT or VF requiring ICD therapy for termination were included in the analysis.
Patients were followed until death from any cause, loss to follow‐up, or December 2015. Information regarding deceased patients was obtained from medical records, family members, patients’ general practitioners, and hospitals to which the patients had been admitted.
2.3. Measurement of fQRS
SAECGs (Predictor BSM‐32, Arrhythmia Research Technology, Fitchburg, Massachusetts, USA) were obtained before and 1 week after CRT implantation. The SAECG was recorded by standard X, Y, Z orthogonal leads using high‐pass filtering. The leads were combined into a vector magnitude (the root of the sum of the squares signals of each lead). Approximately, 200 beats were averaged to obtain a noise level <0.5 μV. The following parameters were calculated: duration of fQRS, low‐amplitude signal duration below 40 μV (LAS 40), and root‐mean‐square voltage in the last 40 ms of the QRS complex (RMS 40).
In this study, narrowing fQRS on an SAECG was defined as a decrease in the fQRS duration from pre‐ to postimplantation of CRT devices.
2.4. Follow‐up echocardiograms
Experienced sonographers performed the echocardiographic studies with a SONOS 5500/iE33/EPIQ7 (Philips Healthcare, Andover, Massachusetts, USA), an ARTIDA (Toshiba Medical Systems, Tochigi, Japan) or a GE Vivid E9 (GE Vingmed Ultrasound AS, Horten, Norway) ultrasound system during continuous electrocardiographic recording before and 6 months after CRT implantation. Echocardiographic data were measured by independent investigators (KA, KA) blinded to the patients’ data.
LV end‐diastolic volume (LVEDV) and LV end‐systolic volume (LVESV) were measured in apical 2‐ and 4‐chamber views. From these results, the LVEF was calculated using the biplane Simpson method. LA volume (LAV) was measured in standard 4‐ and 2‐chamber views. The end‐systolic LAV was measured in each view using the modified Simpson's method. LV mass (LVM) was estimated from the LV cavity dimension and wall thickness at end‐diastole on the M‐mode echocardiogram: LV diastolic dimension (LVDd, cm), interventricular septum thickness (IVS, cm), and LV posterior wall thickness (LVPW, cm). LVM was calculated according to Devereux's formula (Devereux et al., 1986):
The left ventricular mass index (LVMI, g/m2) was defined as the LVM divided by the body surface area (m2). The E‐ and A‐wave maximum velocities and the deceleration time of the E‐wave were measured using pulsed‐wave Doppler of transmitral flow. The E/A ratio was calculated as the ratio of the E‐wave and A‐wave maximum velocities. The E/e′ ration was calculated using the E‐wave maximum velocity and e′ as measured using Doppler tissue imaging of the septal mitral annulus.
Responders to CRT were defined as patients exhibiting a ≥15% reduction in LVESV after 6 months of CRT compared with preimplantation.
2.5. Statistical analysis
The data are presented as the means (SD), medians with ranges or frequencies. The baseline clinical data were compared between groups with narrowing fQRS and nonnarrowing fQRS using Student's t test and the Mann‐Whitney U test. Categorical variables were subjected to chi‐squared analysis. The correlation between ⊿QRS duration on a 12‐lead ECG (postimplantation‐preimplantation) and ⊿fQRS on an SAECG (postimplantation‐preimplantation) was assessed by Pearson's correlation test. The cumulative proportions of the event‐free rate were calculated using the Kaplan‐Meier method. Differences in the event‐free rates were compared using the log‐rank test. Univariate and multivariate analyses using the Cox proportional hazards model were performed to assess the relationships of the following baseline characteristics with mortality: male gender, age ≥65 years, previous sustained VT/VF, LVEF, LBBB before CRT implantation, eGFR <60 ml/min/1.73 m2, responder to CRT, lateral LV lead position, biventricular pacing of >98%, and narrowing fQRS on an SAECG/narrowing QRS on a 12‐lead ECGs. Because narrowing fQRS on an SAECG and narrowing QRS duration on a standard 12‐lead ECGs can be confounding factors, they were analyzed separately. The forward stepwise method was used for the multivariate analyses with entry or removal based on p‐values set at .05. Data analyses were performed with SPSS statistical software (version 11.01, SPSS Inc., Chicago, Illinois, USA).
3. RESULTS
This study included 103 consecutive patients and 53 patients (51%) met the criteria of narrowing fQRS on SAECGs. The baseline clinical characteristics of the patients are summarized in Table 1. Patients with narrowing fQRS were less likely to have previously sustained VT/VF, undergo CRT‐D, and require amiodarone. There were no differences in AV and VV delay or in the distribution of the LV lead position between CRT patients with narrowing fQRS and those with nonnarrowing fQRS. Our patients showed a high biventricular pacing rate and there was no difference in the biventricular pacing rate between CRT patients with narrowing fQRS and those with nonnarrowing fQRS. (98% ± 3% vs 98% ± 3%, p = .160)
Table 1.
Patient characteristics
| Narrowing fQRS(n = 53) | Nonnarrowing fQRS(n = 50) | p‐Value | |
|---|---|---|---|
| Age (years) | 60 ± 14 | 58 ± 13 | .176 |
| Male | 35 (67%) | 39 (78%) | .177 |
| Underlying heart disease | .070 | ||
| Idiopathic dilated cardiomyopathy | 29 (55%) | 24 (48%) | |
| End‐stage hypertrophic cardiomyopathy | 4 (8%) | 13 (26%) | |
| Valvular heart disease | 5 (9%) | 2 (4%) | |
| Others | 15 (28%) | 11 (22%) | |
| Plasma BNP (pg/mL) | 373 (194‐771) | 261 (161‐603) | .254 |
| NYHA functional class | |||
| II/III/IV | 21/30/2 | 25/24/1 | .532 |
| LVEF (%) | 23 ± 8 | 22 ± 7 | .574 |
| eGFR (ml/min/1.73 m2) | 68 ± 31 | 69 ± 38 | .851 |
| Prior sustained VT/VF | 14 (26%) | 23 (46%) | .038 |
| Upgrade to CRT | 21 (40%) | 18 (36%) | .705 |
| CRT‐D | 41 (77%) | 46 (92%) | .040 |
| Medications | |||
| Beta‐blockers | 48 (91%) | 47 (94%) | .515 |
| ACE inhibitors/ARBs | 46 (87%) | 45 (90%) | .612 |
| Mineral corticoid antagonists | 35 (66%) | 34 (68%) | .832 |
| Loop diuretics/thiazides | 34 (64%) | 42 (84%) | .022 |
| Digoxin | 18 (34%) | 22 (44%) | .296 |
| Amiodarone | 19 (36%) | 33 (66%) | .002 |
| Pacing device setting | |||
| AV delay | 167 ± 38 | 172 ± 54 | .473 |
| VV delay | 0 (0–23) | 3 (0–40) | .443 |
| LV lead position | .579 | ||
| Lateral | 26 (49%) | 22 (44%) | |
| Posterior | 0 (0%) | 0 (0%) | |
| Anterior | 16 (31%) | 19 (38%) | |
| Apex | 11 (20%) | 9 (18%) | |
Values are n (%) or means ± SD or median (range).
ACE, angiotensin‐converting enzyme; ARB, angiotensin II receptor blocker; BNP, B‐type natriuretic peptide; CRT, cardiac resynchronization therapy; CRT‐D, CRT with a defibrillator; eGFR, estimated glomerular filtration rate; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; VF, ventricular fibrillation; VT, ventricular tachycardia.
In our patients, the QRS morphological features on 12‐lead ECGs before CRT implantation were distributed as follows: 50 patients had LBBB, 27 patients had non‐LBBB, and 26 patients were right ventricular (RV)‐paced. The rate of narrowing fQRS in patients with a non‐LBBB QRS type was lower than that in those with an LBBB QRS type or RV‐paced QRS type. (37%, 42% and 85%, respectively, p < .001).
3.1. Electrographic parameters
The correlation of QRS reduction as measured by 12‐lead ECGs and SAECGs is demonstrated in Figure 1. Although there was a significant correlation between both, 13 patients showed a narrowing QRS duration on 12‐lead ECGs (defined as a decrease in the QRS duration on a 12‐lead ECG from pre‐ to postimplantation of CRT devices), but not on SAECGs, and seven patients showed the opposite trend.
Figure 1.
Illustration of the correlation of ⊿QRS duration on a 12‐lead ECG (postimplantation‐preimplantation) and ⊿fQRS on an SAECG (postimplantation‐preimplantation)
The QRS duration on 12‐lead ECGs and fQRS on SAECGs before CRT implantation were significantly longer in patients with narrowing fQRS than in patients with nonnarrowing fQRS. However, there were no differences in baseline RMS40 or LAS40 between patients with narrowing fQRS and patients without nonnarrowing fQRS. After CRT implantation, RMS40 was longer and LAS40 was shorter in patients with narrowing fQRS compared with patients with nonnarrowing fQRS. (Table 2).
Table 2.
Electrocardiographic characteristics
| Narrowing f QRS (n = 53) | Nonnarrowing fQRS(n = 50) | p‐Value | |
|---|---|---|---|
| Before implantation of CRT | |||
| 12‐lead ECG | |||
| Heart rate (bpm) | 71 ± 14 | 67 ± 14 | .119 |
| QRS duration (ms) | 187 ± 32 | 145 ± 21 | <.001 |
| QRS type | <.001 | ||
| LBBB | 21 (40%) | 29 (58%) | |
| Non‐LBBB | 10 (19%) | 17 (34%) | |
| RV‐paced | 22 (42%) | 4 (8%) | |
| QT interval (ms) | 457 ± 55 | 450 ± 78 | .484 |
| SAECG | |||
| fQRS | 207 (184–232) | 165 (151–196) | <.001 |
| RMS 40 | 17 (10–31) | 17 (6–34) | .963 |
| LAS 40 | 40 (26–71) | 43 (25–68) | .611 |
| After implantation of CRT | |||
| 12‐lead ECG | |||
| Heart rate (bpm) | 73 ± 13 | 69 ± 10 | .237 |
| QRS duration (ms) | 154 ± 23 | 164 ± 24 | .030 |
| QT interval (ms) | 453 ± 46 | 473 ± 52 | .020 |
| SAECG | |||
| fQRS | 175 (165–196) | 194 (176–231) | .001 |
| RMS 40 | 23 (10–43) | 11 (3–23) | .003 |
| LAS 40 | 35 (26–51) | 47 (36–76) | .001 |
Values are n (%) or means ± SD or median (range).
CRT, cardiac resynchronization therapy; ECG, electrocardiogram; fQRS, filtered QRS duration; LAS 40, low‐amplitude signal duration below 40 μV; LBBB. left bundle branch block; RMS 40, root‐mean‐square voltage in the last 40 ms of the QRS complex; RV, right ventricular; SAECG, signal‐averaged electrocardiogram.
3.2. Relationships between the echocardiographic parameters and narrowing fQRS
During the 6‐month period after implantation, one patient died from congestive HF. The echocardiographic parameters before and 6 months after CRT implantation in 102 patients with narrowing or nonnarrowing fQRS are shown in Table 3. The LVEDV and LVESV were significantly reduced, and the LVEF was significantly increased after CRT in patients with narrowing fQRS, but not in patients with nonnarrowing fQRS. The LVM and LVMI were also significantly reduced after CRT in patients with narrowing fQRS. Although the deceleration time was longer after CRT in both groups, the E/eʹ ratio was significantly reduced only in patients with narrowing fQRS. The LAV and LAVI were significantly reduced in patients with narrowing fQRS, but not in patients with nonnarrowing fQRS.
Table 3.
Echocardiographic changes in patients with narrowing fQRS and patients with nonnarrowing fQRS
| Narrowing fQRS (n = 52) | Nonnarrowing fQRS (n = 50) | |||||
|---|---|---|---|---|---|---|
| Pre CRT | Post CRT | p‐Value | Pre CRT | Post CRT | p‐Value | |
| LVEDV (ml) | 276 ± 126 | 219 ± 88 | <.001 | 243 ± 80 | 223 ± 71 | .090 |
| LVESV (ml) | 216 ± 120 | 155 ± 78 | <.001 | 182 ± 69 | 164 ± 60 | .112 |
| LVEF (%) | 24 ± 8 | 31 ± 7 | <.001 | 26 ± 7 | 28 ± 7 | .264 |
| LVM (g) | 236 ± 93 | 208 ± 80 | .006 | 216 ± 87 | 219 ± 69 | .929 |
| LVMI (g/m2) | 143 ± 51 | 124 ± 46 | .007 | 133 ± 53 | 134 ± 41 | .889 |
| E/A ratio | 1.1 ± 0.7 | 1.0 ± 0.7 | .310 | 1.8 ± 1.6 | 1.3 ± 0.9 | .060 |
| DcT (ms) | 173 ± 72 | 197 ± 46 | .299 | 171 ± 55 | 202 ± 68 | .003 |
| E/eʹ | 21 ± 14 | 15 ± 5 | .310 | 19 ± 8 | 16 ± 8 | .115 |
| LAV (ml) | 88 ± 47 | 71 ± 32 | .018 | 93 ± 33 | 88 ± 42 | .828 |
| LAVI (ml/m2) | 54 ± 28 | 43 ± 20 | .022 | 58 ± 21 | 56 ± 27 | .845 |
| RVSP (mmHg) | 36 ± 12 | 34 ± 11 | .330 | 40 ± 12 | 37 ± 14 | .955 |
Values are means ± SD.
CRT, cardiac resynchronization therapy; DcT, deceleration time; E/A ratio, ratio between E‐wave and A‐wave maximal velocity; E/eʹ, ratio between E‐wave maximal velocity and eʹ‐wave maximal velocity of basal lateral wall; fQRS, filtered QRS duration; LVEDV, left ventricular end‐diastolic volume; LVEDV, left ventricular end‐systolic volume; LVEF, left ventricular ejection fraction; LVM, left ventricular mass; LVMI, left ventricular mass index; LAV, left atrium volume; LAVI, left atrium volume index; RVSP, right ventricular systolic pressure.
Forty‐seven (46%) of 102 patients who underwent the 6‐month follow‐up echocardiography were categorized as responders to CRT. There was a significantly higher proportion of responders to CRT among patients with narrowing fQRS than among patients with nonnarrowing fQRS (58% vs 34%, p = .016).
3.3. Relationships between narrowing fQRS and outcomes
During a median follow‐up period of 33 months, 27 patients (26%) died, 38 patients required hospitalization due to worsened HF, and 39 patients experienced ventricular tachyarrhythmias. Four patients were lost to follow‐up.
Kaplan‐Meier curves for the primary outcome are shown in Figure 2. A significantly lower mortality rate was observed in patients with narrowing fQRS than in patients with nonnarrowing fQRS. Kaplan‐Meier curves for hospitalization due to worsened HF or the occurrence of ventricular tachyarrhythmia are shown in Figure 3. A significantly lower incidence of hospitalization due to worsened HF was observed in patients with narrowing fQRS than in patients with nonnarrowing fQRS. There was a trend toward a lower occurrence of ventricular tachyarrhythmias in patients with narrowing fQRS than in patients with nonnarrowing fQRS. Even in patients who received amiodarone, the differences in primary and secondary outcomes between patients with narrowing fQRS and patients with nonnarrowing fQRS remained the same. (Figures S1 and S2).
Figure 2.
Kaplan‐Meier curves for death from any cause in nonischemic HF patients receiving CRT with narrowing fQRS or nonnarrowing fQRS
Figure 3.
Kaplan‐Meier curves for hospitalization due to worsening HF (a) or the occurrence of ventricular tachyarrhythmias (b) in nonischemic HF patients receiving CRT with narrowing fQRS or nonnarrowing fQRS
In 102 patients who were alive at 6 months after CRT implantation, multivariate analysis revealed that patients with narrowing fQRS had a low risk of the primary outcome (death from any cause) (HR 0.27, 95% CI 0.13–0.70, p = .006), which was independent of eGFR<60 mL/min/1.73 m2 and responder to CRT. (Table 4) Narrowing QRS duration on a 12‐lead ECG was associated with a low risk of the primary outcome in a univariate model, but it was not significant in a multivariate model (HR 0.44, 95% CI 0.19–1.03, p = .059). (Table S1).
Table 4.
Univariate and multivariate analyses for the primary outcome
| Univariate analysis | Multivariate analysis | |||
|---|---|---|---|---|
| HR (95% CI) | p‐Value | HR (95% CI) | p‐Value | |
| Male gender | 0.62 (0.28–1.39) | .246 | ||
| Age ≥ 65 years | 0.88 (0.40–1.94) | .751 | ||
| Previous sustained VT/VF | 1.62 (0.74–3.53) | .228 | ||
| LVEF (per 1% decrease) | 1.04 (0.99–1.09) | .173 | ||
| LBBB before CRT implantation | 0.96 (0.44–2.09) | .921 | ||
| eGFR < 60 ml/min/1.73 m2 | 4.45 (2.00–9.90) | <.001 | 5.71 (2.46–13.30) | <.001 |
| Responder to CRT | 0.33 (0.14–0.77) | .010 | 0.39 (0.16–0.94) | .035 |
| Lateral LV lead | 0.54 (0.22–1.34) | .185 | ||
| Biventricular pacing of >98% | 0.55 (0.19–1.61) | .271 | ||
| Narrowing fQRS on an SAECG | 0.30 (0.13–0.70) | .005 | 0.27 (0.11–0.69) | .006 |
CI, confidence interval; CRT, cardiac resynchronization therapy, eGFR, estimated glomerular filtration rate; fQRS, filtered QRS duration; HR, hazard ratio; LBBB. Left bundle branch block; LV, left ventricular; LVEF, left ventricular ejection fraction; SAECG, signal‐averaged electrocardiogram; VF, ventricular fibrillation; VT, ventricular tachycardia.
4. DISCUSSION
Our study suggested that fQRS narrowing on SAECG predicted outcomes in CRT patients with nonischemic HF: (i) Approximately half of the patients with a newly implanted CRT showed narrowing fQRS on SAECGs. (ii) The LV volume and LVEF significantly improved after CRT in patients with narrowing fQRS, but not in patients with nonnarrowing fQRS, and there was a higher rate of response to CRT in patients with narrowing fQRS compared with patients with nonnarrowing fQRS. (iii) There was a significantly lower mortality rate in patients with narrowing fQRS compared with patients with nonnarrowing fQRS, and narrowing fQRS was associated with a low risk of mortality. (iv) There was a lower incidence of hospitalizations due to worsened HF and a trend toward a lower occurrence of ventricular tachyarrhythmia in patients with narrowing fQRS compared with patients with nonnarrowing fQRS.
Previous reports showed that the proportion of HF patients with a narrowing QRS duration on 12‐lead ECGs after CRT implantation was approximately 60%, including both ischemic and nonischemic etiologies (Molhoek et al., 2004; Rickard et al., 2011; Karaca et al., 2016). In our study, 59 (57%) nonischemic HF patients showed a narrowing QRS duration on 12‐lead ECGs after CRT implantation and 53 (51%) showed narrowing fQRS on SAECGs. Among the study patients, approximately 20% showed different results between changes in the QRS duration on 12‐lead ECGs and changes in fQRS on SAECGs. It is difficult to visually measure the specific QRS duration, and it is also difficult to objectively detect minor changes in the QRS duration using a standard 12‐lead ECG, but fQRS on an SAECG is determined automatically. SAECGs can detect small changes in the QRS duration by averaging a number of ECG cycles. Commercial 12‐lead ECG recording devices provide automatic measurements of the QRS duration. However, a previous study reported that there was low concordance between automated QRS duration measurements and manual QRS duration measurements (De Guillebon et al., 2010). Another study reported that automated QRS durations measured by commercial ECG devices were significantly different from those measured by digitally assisted measurements and that automated QRS duration measurements showed a low predictive value for the CRT response (De Pooter et al., 2015). In addition, an SAECG can detect ventricular late potentials. Therefore, SAECGs provides a better value of the QRS duration compared to 12‐lead ECGs. In our study, the predictive value of narrowing fQRS on SAECGs for the primary outcome was higher than that of narrowing QRS duration on standard 12‐lead ECGs. Therefore, this method may be able to predict the effect of CRT more accurately.
QRS duration has two meanings in HF patients, representing the LV structure and function as well as the presence or absence of an arrhythmogenic substrate. Our results showed that narrowing fQRS on an SAECG in the early phase after CRT implantation could predict both LV reverse remodeling and long‐term outcomes, including the occurrence of ventricular tachyarrhythmias in nonischemic HF patients. Generally, nonischemic HF patients experience a significant benefit from CRT relative to reverse LV remodeling compared to ischemic HF patients, and the magnitude of improvements in LV remodeling and function is high in nonischemic HF patients (Brignole et al., 2013). In nonischemic HF patients receiving CRT, the QRS duration which reflects conduction abnormalities within the QRS complex, is likely to be a predictive marker of improvements in LV structure and function. In our study, improvements in the echocardiographic diastolic parameters and systolic parameters were observed in CRT patients with narrowing fQRS. A substudy of PROSPECT reported diastolic parameters and changes at 6 months after CRT implantation compared with baseline and showed that the E/A ratio and LA area were reduced and appear to be correlated with outcomes at 6 months (Sullivan et al., 2013). This study also showed that the deceleration time increased after CRT, reflecting an improvement from a restrictive filling pattern to impaired relaxation as well as a reduced E/A ratio, although the E/e′ ratio did not show a significant change after CRT (Sullivan et al., 2013). In our study, the LA volume was reduced in CRT patients with narrowing fQRS but not in patients with nonnarrowing fQRS. However, the deceleration time, E/A ratio, and E/e′ ratio did not change in CRT patients with narrowing fQRS. Although it is difficult to evaluate diastolic function in severe systolic failure patients, the improvement of the LA volume in CRT patients with narrowing fQRS, but not in patients with nonnarrowing fQRS, may partially represent improvement in the diastolic function.
Amiodarone has been proven to prolong the QRS duration. In our study, half of patients received amiodarone. However, the use of amiodarone did not influence the main findings of this study. An effect of narrowing fQRS on outcomes was observed in CRT patients despite treatment with amiodarone.
In our study, SAECGs showed that RMS 40 was longer and LAS 40 was shorter after CRT implantation in patients with narrowing fQRS compared with patients with nonnarrowing fQRS, although there were no differences in baseline RMS 40 or LAS 40 between patients. Controversy regarding the benefit of CRT in preventing SCD or ventricular tachyarrhythmias remains. Recently, a meta‐analysis and large cohort study suggested that there was no benefit of additional primary prevention ICD therapy with CRT in HF patients with nonischemic‐dilated cardiomyopathy (Barra et al., 2015, 2017). In our study, a trend toward a lower occurrence of ventricular tachyarrhythmias was observed in patients with narrowing fQRS compared with patients with nonnarrowing fQRS. Patients with narrowing fQRS were less likely to have previously sustained VT/VF and CRT‐D, but a CRT modification on the arrhythmogenic substrate in the LV may affect the occurrence of ventricular arrhythmias. Although the reasons are not fully understood, narrowing fQRS may reflect improved LV systolic function and suppressed arrhythmogenesis in nonischemic HF patients with CRT.
Response to CRT is commonly defined as improvements in echocardiographic parameters or exercise tolerance and NYHA function classes. Response to CRT, which is defined as a reduction of LVESV ≥15% using echocardiography, is known to be a potential predictor of a good prognosis in CRT patients (Ypenburg et al., 2009; Gold et al., 2015). In our study, there was a significantly higher rate of response to CRT in patients with narrowing fQRS than in patients with nonnarrowing fQRS. Recently, renal impairment was also reported to be an independent predictor of long‐term mortality in CRT patients (Daly et al., 2016; Kpaeyeh et al., 2017). However, several factors contribute to the prognosis of HF patients, and the response to CRT alone may not be sufficient for predicting their outcomes. This trend would be the same for AA and AV delay, LV lead position and biventricular pacing rate. It is difficult to predict the outcomes of CRT patients using one surrogate marker, so combined assessments using several markers are required. According to our results, narrowing fQRS on SAECGs may be an additional predictor of outcomes, including mortality and the occurrence of life‐threatening ventricular arrhythmias in CRT patients. To confirm this issue, a further evaluation is needed.
4.1. Study limitations
There were some limitations in this study. First, this study was a retrospective observational study in a single center. Data concerning clinical conditions at the time of death, cardiovascular events and ICD therapy were not available. In addition, there was a treatment bias. Second, the number of subjects was relatively small. Therefore, subgroup analyses were not feasible. Third, ICD detection programmed for the VF zone and VT zone and ICD therapy settings including antitachycardia pacing and shock were not identical.
5. CONCLUSIONS
Our results suggested that narrowing fQRS on SAECGs after CRT implantation predicts LV reverse remodeling, long‐term outcomes and ventricular arrhythmias in nonischemic HF patients.
CONFLICTS OF INTEREST
Dr. Shoda is a professor of an endowed department that is supported with unrestricted grants from Medtronic, Boston‐Scientific, Biotronik, and St Jude Medical. The other authors indicated no potential conflicts of interest.
COMPETING INTERESTS
None declared.
Supporting information
ACKNOWLEDGMENTS
We thank Mr. Koichi Takeuchi for his assistance.
Suzuki A, Shiga T, Yagishita D, et al. Narrowing filtered QRS duration on signal‐averaged electrocardiogram predicts outcomes in cardiac resynchronization therapy patients with nonischemic heart failure. Ann Noninvasive Electrocardiol. 2018;23:e12523 10.1111/anec.12523
Funding informationThis study was supported by a grant from the Health and Labour Sciences Research Grant H25‐Junkankito (Seishuu‐Ippan‐018)
REFERENCES
- Andrikopoulos, G. K. , Tzeis, S. , Kolb, C. , Sakellariou, D. , Avramides, D. , Alexopoulos, E. C. , … Manolis, A. S. (2009). Correlation of mechanical dyssynchrony with QRS duration measured by signal‐averaged electrocardiography. Annals of Noninvasive Electrocardiology, 14, 234–241. 10.1111/j.1542-474x.2009.00303.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Barra, S. , Boveda, S. , Providência, R. , Sadoul, N. , Duehmke, R. , Reitan, C. , … French‐UK‐Sweden CRT Network . (2017). Adding defibrillation therapy to cardiac resynchronization on the basis of the myocardial substrate. Journal of the American College of Cardiology, 69, 1669–1678. 10.1016/j.jacc.2017.01.042 [DOI] [PubMed] [Google Scholar]
- Barra, S. , Providência, R. , Tang, A. , Heck, P. , Virdee, M. , & Agarwal, S. (2015). Importance of Implantable Cardioverter‐Defibrillator Back‐Up in Cardiac Resynchronization Therapy Recipients: A Systematic Review and Meta‐Analysis. Journal of the American Heart Association, 4, e002539 10.1161/jaha.115.002539 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brignole, M. , Auricchio, A. , Baron‐Esquivias, G. , Bordachar, P. , Boriani, G. , Breithardt, O. A. , … Vardas, P. E. (2013). 2013 ESC guidelines on cardiac pacing and cardiac resynchronization therapy: The task force on cardiac pacing and resynchronization therapy of the European Society of Cardiology (ESC). Developed in collaboration with the European Heart Rhythm Association (EHRA). EP Europace, 15, 1070–1118. 10.1093/europace/eut206 [DOI] [PubMed] [Google Scholar]
- Daly, D. D. Jr , Maran, A. , Hyer, J. M. , Funke, F. , Waring, A. , Cuoco, F. A. , … Gold, M. R. (2016). The Effect of Chronic Kidney Disease on Mortality with Cardiac Resynchronization Therapy. Pacing and Clinical Electrophysiology, 39, 863–869. 10.1111/pace.12883 [DOI] [PMC free article] [PubMed] [Google Scholar]
- De Guillebon, M. , Thambo, J. B. , Ploux, S. , Deplagne, A. , Sacher, F. , Jais, P. , … Bordachar, P. (2010). Reliability and reproducibility of QRS duration in the selection of candidates for cardiac resynchronization therapy. Journal of Cardiovascular Electrophysiology, 21, 890–892. 10.1111/j.1540-8167.2010.01743.x [DOI] [PubMed] [Google Scholar]
- De Pooter, J. , El Haddad, M. , Timmers, L. , Van Heuverswyn, F. , Jordaens, L. , Duytschaever, M. , & Stroobandt, R. (2015). Different Methods to Measure QRS Duration in CRT Patients: Impact on the Predictive Value of QRS Duration Parameters. Annals of Noninvasive Electrocardiology, 21, 305–315. 10.1111/anec.12313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Devereux, R. D. , Alonso, D. R. , Lutas, E. M. , Gottileb, G. J. , Campo, E. , Sachs, I. , & Reichek, N. (1986). Echocardiographic assessment of left ventricular hypertrophy: Comparison to necropsy findings. The American Journal of Cardiology, 57, 450–458. 10.1016/0002-9149(86)90771-x [DOI] [PubMed] [Google Scholar]
- Gold, M. R. , Daubert, C. , Abraham, W. T. , Ghio, S. , St John Sutton, M. , Hudnall, J. H. , … Linde, C. (2015). The effect of reverse remodeling on long‐term survival in mildly symptomatic patients with heart failure receiving cardiac resynchronization therapy: Results of the REVERSE study. Heart Rhythm: the Official Journal of the Heart Rhythm Society, 12, 524–530. 10.1016/j.hrthm.2014.11.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Goldberger, J. J. , Cain, M. E. , Hohnloser, S. H. , Kadish, A. H. , Knight, B. P. , Lauer, M. S. , … Zipes, D. P. (2008). American Heart Association/American College of Cardiology Foundation/Heart Rhythm Society scientific statement on noninvasive risk stratification techniques for identifying patients at risk for sudden cardiac death: A scientific statement from the American Heart Association Council on Clinical Cardiology Committee on Electrocardiography and Arrhythmias and Council on Epidemiology and Prevention. Circulation, 118, 1497–1518. 10.1161/circulationaha.107.189375 [DOI] [PubMed] [Google Scholar]
- Hsing, J. M. , Selzman, K. A. , Leclercq, C. , Pires, L. A. , McLaughlin, M. G. , McRae, S. E. , … Zimetbaum, P. J. (2011). Paced left ventricular QRS width and ECG parameters predict outcomes after cardiac resynchronization therapy: PROSPECT‐ECG substudy. Circulation: Arrhythmia and Electrophysiology, 4, 851–857. 10.1161/circep.111.962605 [DOI] [PubMed] [Google Scholar]
- JCS Joint Working Group (2013). Guidelines for Non‐Pharmacotherapy of Cardiac Arrhythmias (JCS 2011) –Digest Version‐. Circulation Journal: Official Journal of the Japanese Circulation Society, 77, 249–274. 10.1253/circj.cj-66-0054 [DOI] [PubMed] [Google Scholar]
- Karaca, O. , Omaygenc, M. O. , Cakal, B. , Cakal, S. D. , Gunes, H. M. , Barutcu, I. , … Kilicaslan, F. (2016). Effect of QRS Narrowing After Cardiac Resynchronization Therapy on Functional Mitral Regurgitation in Patients With Systolic Heart Failure. The American Journal of Cardiology, 117, 412–419. 10.1016/j.amjcard.2015.11.010 [DOI] [PubMed] [Google Scholar]
- Kpaeyeh, J. A. Jr , Divoky, L. , Hyer, J. M. , Daly, D. D. Jr , Maran, A. , Waring, A. , & Gold, M. R. (2017). Impact of Renal Function on Survival After Cardiac Resynchronization Therapy. The American Journal of Cardiology, 120, 262–266. 10.1016/j.amjcard.2017.04.017 [DOI] [PubMed] [Google Scholar]
- Lecoq, G. , Leclercq, C. , Leray, E. , Crocq, C. , Alonso, C. , de Place, C. , … Daubert, C. (2005). Clinical and electrocardiographic predictors of a positive response to cardiac resynchronization therapy in advanced heart failure. European Heart Journal, 26, 1094–1100. 10.1093/eurheartj/ehi146 [DOI] [PubMed] [Google Scholar]
- Mancini, D. M. , Wong, K. L. , & Simson, M. B. (1993). Prognostic value of an abnormal signal‐averaged electrocardiogram in patients with nonischemic congestive cardiomyopathy. Circulation, 87, 1083–1092. 10.1161/01.cir.87.4.1083 [DOI] [PubMed] [Google Scholar]
- Mohajer, K. , Cheng, J. , Siu, J. , Michael, K. A. , Abdollah, H. , Simpson, C. S. , … Pedfearn, D. P. (2014). Signal‐averaged versus 12‐lead electrocardiography in measuring QRS duration for cardiac resynchronization therapy: The role of ventricular late potentials. The Journal of Innovations in Cardiac Rhythm Management, 5, 1562–1567. [Google Scholar]
- Molhoek, S. G. , Van Erven, L. , Bootsma, M. , Steendijk, P. , Van Der Wall, E. E. , & Schalij, M. J. (2004). QRS duration and shortening to predict clinical response to cardiac resynchronization therapy in patients with end‐stage heart failure. Pacing and Clinical Electrophysiology, 27, 308–313. 10.1111/j.1540-8159.2004.00433.x [DOI] [PubMed] [Google Scholar]
- Rickard, J. , Cheng, A. , Spragg, D. , Cantillon, D. , Chung, M. K. , Tang, W. H. , … Varma, N. (2013). QRS narrowing is associated with reverse remodeling in patients with chronic right ventricular pacing upgraded to cardiac resynchronization therapy. Heart Rhythm: the Official Journal of the Heart Rhythm Society, 10, 55–60. 10.1016/j.hrthm.2012.09.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rickard, J. , Popovic, Z. , Verhaert, D. , Sraow, D. , Baranowski, B. , Martin, D. O. , … Chung, M. K. (2011). The QRS narrowing index predicts reverse left ventricular remodeling following cardiac resynchronization therapy. Pacing and Clinical Electrophysiology, 34, 604–611. 10.1111/j.1540-8159.2010.03022.x [DOI] [PubMed] [Google Scholar]
- Sullivan, R. M. , Murillo, J. , Gerritse, B. , Chung, E. , Orlov, M. V. , Stegemann, B. , … Olshansky, B. (2013). Do baseline diastolic echocardiographic parameters predict outcome after resynchronization therapy? Results from the PROSPECT trial. Pacing and Clinical Electrophysiology, 36, 214–220. 10.1111/pace.12042 [DOI] [PubMed] [Google Scholar]
- Ypenburg, C. , van Bommel, R. J. , Borleffs, C. J. , Bleeker, G. B. , Boersma, E. , Schalij, M. J. , & Bax, J. J. (2009). Long‐term prognosis after cardiac resynchronization therapy is related to the extent of left ventricular reverse remodeling at midterm follow‐up. Journal of the American College of Cardiology, 53, 483–490. 10.1016/j.jacc.2008.10.032 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials