OUP user menu

New-onset left bundle branch block independently predicts long-term mortality in patients with idiopathic dilated cardiomyopathy: data from the Trieste Heart Muscle Disease Registry

Aneta Aleksova, Cosimo Carriere, Massimo Zecchin, Giulia Barbati, Giancarlo Vitrella, Andrea Di Lenarda, Gianfranco Sinagra
DOI: http://dx.doi.org/10.1093/europace/euu016 1450-1459 First published online: 18 February 2014

Abstract

Aims Left bundle branch block (LBBB) is commonly associated with heart failure. We evaluated the prevalence, incidence, and impact of LBBB on long-term outcome in young patients with heart failure affected by idiopathic dilated cardiomyopathy (DCM).

Methods and results We included 608 patients with DCM from the Heart Muscle Disease Registry of Trieste in this retrospective analysis. At baseline electrocardiogram (ECG), 189 patients (31.1%) had LBBB. The patients with baseline LBBB had a significantly higher mortality rate than the patients without LBBB (38.6 vs. 27.9%, P = 0.002) at the univariate analysis. After a multiple covariate adjustment, the baseline LBBB was not associated with a significantly increased risk of death [hazard ratio (HR) 1.27, 95% confidence interval (CI): 0.88–1.81, P = 0.2]. Forty-seven (11.2%) patients without LBBB at baseline ECG developed LBBB during follow-up. Among these, the mortality rate was 49 vs. 25% in patients without new-onset LBBB (P = 0.001). New-onset LBBB was a strong and independent predictor of all-cause mortality (HR 3.18, 95% CI: 1.90–5.31, P < 0.001) at multivariate analysis.

Conclusion After correcting for potential confounders, new-onset LBBB was found to be associated with an increased risk of all-cause mortality. The management of patients with new-onset LBBB may need to be more aggressive, possibly including early cardiac resynchronization therapy/implantable cardioverter-defibrillator therapy.

  • Idiopathic dilated cardiomyopathy
  • Heart failure
  • Left bundle branch block
  • Prognosis

What's new?

  • Our study provides an extensive evaluation of the influence of left bundle branch block (LBBB) on long-term prognosis in a large, single-centre, well-characterized population of young patients affected by idiopathic dilated cardiomyopathy. Data from our study suggest that the development of LBBB in patients who are already on treatment with angiotensin-converting enzyme inhibitors and beta-blockers is associated with a worse prognosis. Patients developing new-onset LBBB during follow-up may benefit from a tighter follow-up and a more aggressive management.

Introduction

Idiopathic dilated cardiomyopathy (DCM) is one of the most common causes of heart failure (HF) and heart transplantation (HTx).1 Although in recent years, the outcome of these patients has been improving, mortality is still high.24 Therefore, an identification of valid prognostic factors is mandatory.

Cardiac resynchronization therapy (CRT) has beneficial effects in patients with HF.57 Left bundle branch block (LBBB), a frequent finding in patients with HF810 may identify patients who benefit the most from CRT.11,12

However, the published data exploring its impact on the outcome in patients with HF are not conclusive.810,1315

In this retrospective study, we evaluated the prevalence of LBBB, the incidence of new-onset LBBB, and their impact on the long-term outcome in a large, single-centre, well-characterized population of young patients with DCM on tailored medical therapy.

Methods

Study population

Six hundred and eight consecutive patients with DCM, enroled in the Heart Muscle Disease Registry of Trieste between January 1998 and December 2007, were included in this analysis.

At the time of enrolment, all patients underwent a complete clinical evaluation, which included an accurate clinical history and a physical examination, blood sampling for laboratory tests, a 12-lead electrocardiogram (ECG), standard X-ray, 24 h Holter monitoring, mono, two-dimensional, and Doppler echocardiography, exercise stress testing, and coronary angiography. Endomyocardial biopsy was performed routinely in all patients before 1996. Thereafter, biopsy was performed only in patients with recent-onset HF and/or clinical history suggesting active myocarditis: (i) recent-onset HF (within 6 months from enrolment), and (ii) severe left ventricular (LV) dysfunction [LV ejection fraction (LVEF) <40%] without ventricular remodelling on echocardiography, and/or (iii) idiopathic major ventricular arrhythmias [MVAs; sustained ventricular tachycardias/ventricular flutter or fibrillation, aborted sudden death (SD)].16 All bioptic samples were tested with an immunohistochemical analysis using specific antibodies for the identification of myocardial inflammation and for the identification, localization, and characterization of mononuclear cell infiltrates (Major histocompatibility complex, class II, DR; CD54 for adhesion molecules; CD4 for helper T-cells; CD8 for suppressor T-cells; CD25 for regulatory T-cells; CD2 for natural killer lymphocytes; CD45RO for memory T-cells; and CD68-PGM1 for histiocytes).16 Diagnosis of acute myocarditis was made in accordance with the Dallas Criteria,17 and confirmed when an immunohistochemical analysis detected a myocardial immune activation in addition to the positivity of the Dallas Criteria. After the year 2000, the bioptic samples were systematically evaluated for the presence of the genome of cardiotropic viruses (parvovirus B19, adenovirus, enterovirus, Epstein–Barr virus, herpes simplex virus 1, herpes simplex virus 2) by real-time polymerase chain reaction using specific primers and probes; in case of a virus-positive endomyocardial biopsy, blood samples were also tested for the same virus.16,18

Dilated cardiomyopathy was diagnosed19 when the patients were found to have LV systolic dysfunction (LVEF < 50%) in the absence of any other known cardiac disease. The patients were excluded if LV dysfunction was secondary to one of the following: hypertension (>160/100 mmHg), significant coronary artery disease history of alcohol abuse (>100 g alcohol/day), tachycardia-induced cardiomyopathy, Cor pulmonale, diseases of pericardium, or congenital heart diseases.

All patients immediately started treatment with angiotensin-converting enzyme (ACE) inhibitors and diuretics. Digoxin was associated when considered necessary. After stabilization, beta-blockers were added at the highest tolerated dosage. Starting in 1998, selected high-risk patients [defined by persistent severe systolic dysfunction and New York Heart Association (NYHA) Class II or III despite optimal medical therapy] underwent implantable cardioverter-defibrillator (ICD) placement in primary prevention, or CRT defibrillator (CRT-D), if patients also had complete LBBB.

Follow-up visits were performed at 6, 12, and 24 months and subsequently every 2 years or more frequently if clinically indicated. At each follow-up visit, all patients underwent a physical examination, a 12-lead ECG, 24 h ECG-Holter monitoring, echocardiogram, and a routine laboratory testing.

Definitions and endpoints

Left bundle branch block was deemed to be present if QRS duration was ≥120 ms and the established morphological criteria were present.20 New-onset LBBB was defined as the development of LBBB during follow-up, in patients who did not have the conduction abnormality at baseline ECG. All ECGs were carefully reviewed by two different cardiologists (blinded to the clinical status of patients) to properly assess the LBBB criteria.

We analysed the impact of LBBB at baseline or new-onset LBBB during follow-up on all-cause mortality/urgent HTx in Status I, SD and MVAs, or death from HF.

Heart transplantation was performed in patients with refractory HF needing continuous intravenous inotropic treatment or mechanical circulatory support (Status I). Sudden death was defined ‘unexpected’ when it occurred within 1 h after the beginning of symptoms or during sleep in stable patients in NYHA II–III class. The following were considered MVAs: ventricular fibrillation/flutter, sustained ventricular tachycardias, and appropriate shock of ICD in case of ventricular tachycardia/ventricular fibrillation with ventricular rate >220/min. Death occurring during progressive HF was considered as pump failure death.

Information regarding the endpoints was obtained directly from patients, from their family doctor, and from death registries of the communes of residence.

The end of follow-up was set to 31 August 2010 or the date of death or HTx of the patient. This study conforms to the principles outlined in the Declaration of Helsinki and was approved by the institutional ethics committee. All patients provided written informed consent prior to enlistment in the Registry.

Statistical analysis

Continuous variables are reported as mean ± standard deviation, and the nominal data as counts and percentages. Continuous variables were compared across the groups by analysis of variance (ANOVA), using the Brown–Forsythe statistic when the assumption of equal variances did not hold. The χ2 test was used to compare binary variables, with Yates' correction for continuity when necessary. Event-free survival curves were plotted with the Kaplan–Meier method, and the log-rank test for differences in survival was applied to assess the relationship between baseline/new-onset LBBB and long-term outcome.

Cox proportional hazards multivariate model was used to evaluate the relationship between baseline data and long-term outcome. Covariates were selected from univariable Cox regression models performed on all the parameters recorded at baseline (data not shown) including all the variables with a P ≤ 0.1 as well as clinically relevant variables. All variables were inserted in a full initial multivariable model and reduced by means of a backward-conditional stepwise procedure to minimize the collinearity among the predictors. The same model was fitted for the patients without LBBB at baseline by a substitution of the baseline LBBB with new-onset LBBB. A proportional hazard assumption for this variable was checked by plotting the corresponding Schoenfeld residuals against a fitted time and with the Grambsch and Therneau test.21 Then, the new-onset LBBB was considered as a time-dependent covariate in a Cox regression analysis. The effect of the new-onset LBBB on the outcome was visually represented by a ‘landmark-type’ analysis. The follow-up time was divided into several periods starting from 1 year of follow-up. Patient survival was then described with the estimated Cox multivariable model conditional on the patient being alive at the start of the period, and having developed an LBBB prior to the end of each interval. To have sufficient power for a hazard ratio estimation in each period, all the events occurring after the start of the period were included in the endpoint. This approach provides a general visual trend of the adjusted association between the cumulative effect of the independent variable (new-onset LBBB) and the dependent variable (death/urgent HTx) over time.

While considering all the parameters measured at baseline, univariable and multivariable logistic regression models were also estimated to find the possible predictors of new-onset LBBB. However, we did not find any statistically significant parameters or a combination of parameters to be predictive of this condition.

To exclude the impact of CRT and ICD treatments on the prognostic value of LBBB in patients with DCM, we also performed subgroup survival analyses on a reduced sample size of 536 patients who did not undergo device implantation during follow-up.

A two-tailed P ≤ 0.05 was considered statistically significant for all the test results.

All the analyses were performed by using SPSS Statistical Package for Windows, Release 13.0 and R statistical package version 2.7.2.

Results

Patient population

Baseline characteristics of our patient population are summarized in Table 1. The mean (SD) age was 46 ± 13 years; 74% were males and 25% of them were in NYHA Class III–IV. Left bundle branch block was present in 189 (31.1%) patients. An unadjusted comparison revealed that patients with baseline LBBB were likely to be older, to have a longer duration of HF symptoms, and to have a poorer renal function. At an echocardiographic evaluation, the patients with LBBB at enrolment had a more dilated left ventricle, and a lower LVEF. The pharmacological treatment did not differ significantly between the groups.

View this table:
Table 1

Baseline characteristics of the study patients by the presence of LBBB

All patients (n = 608)LBBB
(n = 189, 31.1%)
No LBBB (n = 419, 68.9%)P valuea
Age (years)46 ± 1351 ± 1143 ± 13<0.001
Male gender (%)7469760.05
BMI (kg/m2)26 ± 426 ± 426 ± 40.34
Familial history of DCM (%)20825<0.001
HF duration (months)12 ± 2418 ± 329 ± 19<0.001
SBP (mmHg)126 ± 17128 ± 17125 ± 160.04
DBP (mmHg)80 ± 1181 ± 1180 ± 110.17
Heart rate (b.p.m.)78 ± 1576 ± 1379 ± 150.02
NYHA III–IV (%)2528240.3
Signs of congestion (%)1215110.25
Bilateral ankle oedema (%)5560.67
AF total (%)1511160.1
RBBB (%)40.550.005
Hb (g/dL)14 ± 214 ± 114 ± 20.39
Na+ (mEq/L)141 ± 3140 ± 3141 ± 40.54
Creatinine (mg/dL)1 ± 0.21 ± 0.21 ± 0.20.89
GFR (mL/min)90 ± 2682 ± 2493 ± 27<0.001
CKD (%)111590.05
Anaemia (%)98100.48
LVEDD (mm)67 ± 969 ± 1066 ± 90.007
Indexed LVEDD (mm/m2)36 ± 537 ± 535 ± 5<0.001
LVESD (mm)56 ± 1158 ± 1155 ± 10<0.001
Indexed LVESD (mm/m2)30 ± 631 ± 629 ± 6<0.001
LVEDV (mL)193 ± 76211 ± 86185 ± 69<0.001
Indexed LVEDV (mL/m2)102 ± 38114 ± 4497 ± 34<0.001
LVESV (mL)138 ± 67158 ± 76128 ± 60<0.001
Indexed LVESV (mL/m2)73 ± 3585 ± 4067 ± 30<0.001
LVEF (%)31 ± 1027 ± 933 ± 9<0.001
LVEF < 30% (%)506642<0.001
RV FAC (%)44 ± 1651 ± 1642 ± 16<0.001
RVH (%)0.400.60.47
LA area (cm2)26 ± 926 ± 926 ± 90.77
Indexed LA area (cm2)14 ± 514 ± 514 ± 50.31
LA diameter (mm)39 ± 840 ± 839 ± 80.3
Indexed LA diameter (mm/m2)21 ± 422 ± 521 ± 40.03
Restrictive filling pattern (%)2726280.64
Moderate–severe MR (%)b6154640.01
MAP (mmHg)87 ± 1387 ± 1287 ± 130.87
CI (L/min/m2)3.5 ± 13.5 ± 13.5 ± 10.98
CO (L/min)6.6 ± 26.5 ± 26.6 ± 20.69
RAP (mmHg)5 ± 35 ± 35 ± 30.83
PCWP (mmHg)13 ± 811 ± 613 ± 80.02
ACE inhibitors (%)9091890.46
ARBs (%)9990.95
Beta-blockers (%)8386820.24
Diuretics (%)6165590.19
Aldosterone receptor antagonists (%)3032290.51
Nitrates (%)81170.12
Amiodarone (%)2018220.16
Digoxin (%)6668650.45
Antiplatelet agents (%)1620140.1
Anticoagulants (%)1915210.07
ICD during follow-up (%)1113100.33
CRT during follow-up (%)4830.006
  • ACE, angiotensin-converting enzyme; AF, atrial fibrillation; ARBs, angiotensin receptors blockers; BMI, body mass index; CI, cardiac index; CO, cardiac output; CKD, chronic kidney disease; CRT, cardiac resynchronization therapy; DBP, diastolic blood pressure; DCM, dilated cardiomyopathy; FAC, fractional area change; GFR, glomerular filtration rate; Hb, haemoglobin; HF, heart failure; ICD, implantable cardioverter-defibrillator; LA, left atrium; LBBB, left bundle branch block; LVEDD, left ventricular end-diastolic diameter; LVEDV, left ventricular end-diastolic volume; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic diameter; LVESV, left ventricular end-systolic volume; MAP, mean arterial pressure; MR, mitral regurgitation; NYHA, New York Heart Association functional stage; PCWP, pulmonary capillary wedge pressure; RAP, right atrial pressure; RBBB, right bundle branch block; RV, right ventricle; RVH, right ventricular hypertrophy; SBP, systolic blood pressure.

  • *Values are reported as mean ± standard deviation or as %.

  • aP value from the ANOVA test for continuous variables and the Pearson χ2 tests for categorical variables.

  • bMitral regurgitation defined as: moderate regurgitation (jet area 4–8 cm2); severe regurgitation (jet area > 8 cm2).

Baseline left bundle branch block and outcomes

During a median follow-up of 122 (interquartile range 65–180) months, 191 patients (31.4%) died or underwent urgent HTx (141 patients died; 50 patients underwent urgent HTx). At the Kaplan–Meier univariable analysis, the total mortality/urgent HTx rate was significantly higher in patients with baseline LBBB than in those free of LBBB at the baseline (38.6 vs. 27.9%, P = 0.002; Figure 1A). In a further survival analysis, the baseline LBBB was associated with an increased incidence of SD/MVAs (18 vs. 12.6%, P = 0.035; Figure 1B), but not with circulatory failure death (14.8 vs. 13%, P = 0.278; Figure 1C).

Figure 1

Kaplan–Meier curves comparing the survival free of death/urgent HTx (A), SD/MVAs (B), and death for HF (C) in patients with DCM presenting LBBB or not at the baseline. D/Htx, death/urgent heart transplantation; DCM, dilated cardiomyopathy; DHF, death for heart failure; LBBB, left bundle branch block; pts, patients; SD/MVA, sudden death/major ventricular arrhythmias.

After an adjustment for baseline covariates in the Cox multivariate analysis, the presence of LBBB at the baseline was no longer significantly associated with all-cause mortality [hazard ratio (HR) 1.27, 95% confidence interval (CI): 0.88–1.81, P = 0.2]. An almost identical HR for baseline LBBB (HR 1.13, 95% CI 0.79–1.63, P = 0.3) was found when the analyses were repeated in the restricted study population of 536 patients who did not undergo device implantation during follow-up (for details of case selection, see the Statistical analysis section).

The variables associated with mortality during follow-up included gender, NYHA class, systolic blood pressure, duration of HF, LVEF, the presence of ICD/CRT, and the use of beta-blockers (Table 2).

View this table:
Table 2

Multivariable analysis of risk of all-cause mortality/urgent HTx

HR95% CIP value
Gender (male vs. female)1.831.25–2.680.002
SBP (for 10 mmHg decrease)1.131.02–1.240.01
NYHA (III–IV vs. I–II)1.431.003–2.0310.048
LVEF (for 10 points % decrease)1.571.29–1.91<0.001
Duration of HF (for 24 months increase)1.301.19–1.43<0.001
Beta-blockers (yes vs. no)0.530.37–0.76<0.001
ICD/CRT (yes vs. no)0.620.38–0.980.05
  • CI, confidential interval; CRT, cardiac resynchronization therapy; HF, heart failure; HR, hazard ratio; ICD, implantable cardioverter-defibrillator; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association functional stage; SBP, systolic blood pressure.

Incidence of new-onset left bundle branch block during the study

New-onset LBBB occurred during follow-up in 47 (11.2%) of 419 patients without LBBB at the baseline. There were no differences in baseline clinical parameters between the patients who developed LBBB and those who did not (Table 3). Digoxin treatment was more frequent among the patients with new-onset LBBB. We did not find any baseline clinical parameter to be predictive of new-onset LBBB.

View this table:
Table 3

Baseline characteristics of patients without LBBB at baseline stratified by the LBBB development during the follow-up

All patients
(n = 419)
LBBB follow-up (n = 47, 11.2%)No LBBB follow-up (n = 372, 88.8%)P valuea
Age (years)43 ± 1347 ± 1443 ± 130.06
Male gender (%)7674770.74
BMI (kg/m2)26 ± 426 ± 426 ± 40.96
Familial history of DCM (%)2527250.66
HF duration (months)9 ± 1912 ± 239 ± 180.27
SBP (mmHg)125 ± 16125 ± 17125 ± 160.85
DBP (mmHg)80 ± 1179 ± 980 ± 110.69
Heart rate (b.p.m.)79 ± 1579 ± 1479 ± 160.9
NYHA III–IV (%)2428240.54
Signs of congestion (%)1111110.85
Bilateral ankle oedema (%)6460.64
AF total (%)1619160.57
RBBB (%)5260.32
Hb (g/dL)14 ± 214 ± 114 ± 20.89
Na+ (mEq/L)141 ± 4141 ± 4141 ± 40.34
Creatinine (mg/dL)1 ± 0.21 ± 0.21 ± 0.20.43
GFR (mL/min)93 ± 2790 ± 3094 ± 260.39
CKD (%)91090.95
Anaemia (%)101690.25
LVEDD (mm)66 ± 969 ± 1166 ± 80.07
Indexed LVEDD (mm/m2)35 ± 536 ± 535 ± 50.23
LVESD (mm)55 ± 1057 ± 1255 ± 100.16
Indexed LVESD (mm/m2)29 ± 630 ± 629 ± 60.36
LVEDV (mL)185 ± 69207 ± 95182 ± 650.03
Indexed LVEDV (mL/m2)97 ± 34106 ± 4296 ± 330.06
LVESV (mL)128 ± 60146 ± 79126 ± 570.04
Indexed LVESV (mL/m2)67 ± 3074 ± 3766 ± 290.09
LVEF (%)32 ± 1032 ± 1033 ± 100.58
LVEF < 30% (%)4245410.63
RV FAC (%)42 ± 1646 ± 1641 ± 160.14
RVH (%)0.600.80.62
LA area (cm2)26 ± 925 ± 826 ± 90.44
Indexed LA area (cm2)14 ± 513 ± 414 ± 50.27
LA diameter (mm)39 ± 839 ± 939 ± 80.87
Indexed LA diameter (mm/m2)21 ± 420 ± 421 ± 40.63
Restrictive filling pattern (%)2831280.69
Moderate–severe MR (%)b6465640.92
MAP (mmHg)87 ± 1388 ± 1487 ± 130.6
CI (L/min/m2)3.5 ± 13.6 ± 13.5 ± 10.56
CO (L/min)6.6 ± 27 ± 26.6 ± 20.34
RAP (mmHg)5 ± 35 ± 35 ± 40.61
PCWP (mmHg)13 ± 813 ± 913 ± 80.93
ACE inhibitors (%)8998880.04
ARBs (%)9490.24
Beta-blockers (%)8285810.54
Diuretics (%)5968580.2
Aldosterone receptor antagonists (%)3036290.3
Nitrates (%)7870.65
Amiodarone (%)2223220.87
Digoxin (%)6581630.02
Antiplatelet agents (%)148150.23
Anticoagulant (%)2123210.67
ICD during follow-up (%)10288<0.001
CRT during follow-up (%)2171<0.001
  • ACE, angiotensin-converting enzyme; AF, atrial fibrillation; ARBs, angiotensin receptors blockers; BMI, body mass index; CI, cardiac index; CO, cardiac output; CKD, chronic kidney disease; DBP, diastolic blood pressure; DCM, dilated cardiomyopathy; FAC, fractional area change; GFR, glomerular filtration rate; Hb, haemoglobin; HF, heart failure; LA, left atrium; LBBB, left bundle branch block; LVEDD, left ventricular end-diastolic diameter; LVEDV, left ventricular end-diastolic volume; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic diameter; LVESV, left ventricular end-systolic volume; MAP, mean arterial pressure; MR, mitral regurgitation; NYHA, New York Heart Association functional stage; PCWP, pulmonary capillary wedge pressure; RAP, right atrial pressure; RBBB, right bundle branch block; RV, right ventricle; RVH, right ventricular hypertrophy; SBP, systolic blood pressure.

  • *Values are reported as mean ± standard deviation or as %.

  • aP value from the ANOVA test for continuous variables and the Pearson χ2 tests for categorical variables.

  • bMitral regurgitation defined as: moderate regurgitation (jet area 4–8 cm2); severe regurgitation (jet area > 8 cm2).

In 22 of 47 patients, new-onset LBBB occurred during the 1 year of follow-up. This subgroup was more likely to have moderate-to-severe mitral insufficiency (53 vs. 21%, P = 0.03), a poorer renal function (78 ± 19 vs. 99 ± 35 mL/min, P = 0.03), a longer history of HF (20 ± 31 vs. 4 ± 7 months, P = 0.015), and was more frequently on diuretic treatment (91 vs. 48%, P = 0.002) compared with the patients who developed LBBB later in follow-up.

New-onset left bundle branch block and outcomes

During follow-up, 191 (31.4%) patients with DCM died or underwent urgent HTx. The unadjusted total mortality/HTx risk was significantly higher (49 vs. 25%, P = 0.001) in the patients with new-onset LBBB compared with the patients without new-onset LBBB. Kaplan–Meier mortality curve is displayed in Figure 2A. The patients with new LBBB experienced a higher rate of SD/MVAs (30 vs. 11%, P < 0.001; Figure 2B) and mortality caused by pump failure (11 vs. 4%, P = 0.034; Figure 2C). New-onset LBBB treated as a time-dependent variable was confirmed as a predictor of outcome (HR 3.38, 95% CI: 2.12–5.39, P < 0.001) at the Cox univariable analysis.

Figure 2

Kaplan–Meier curves comparing the survival free of death/urgent HTx (A), SD/MVAs (B), and death for HF (C) in patients with DCM developing LBBB or not during the long-term follow-up. D/Htx, death/urgent heart transplantation; DCM, dilated cardiomyopathy; DHF, death for heart failure; LBBB, left bundle branch block; pts, patients; SD/MVA, sudden death/major ventricular arrhythmias.

Furthermore, at multivariate analysis, new-onset LBBB was an independent predictor of the subsequent all-cause mortality when treated as a time-dependent variable (HR 3.18, 95% CI: 1.90–5.31, P < 0.001) (Table 4), if it occurred after the 1 year of follow-up (Figure 3).

View this table:
Table 4

Multivariate analysis: the role of the new-onset LBBB on total mortality/urgent HTx rate

HR95% CIP value
Gender (male vs. female)1.681.01–2.800.04
SBP (for 10 mmHg decrease)1.221.06–1.400.01
NYHA (III–IV vs. I–II)1.500.97–2.330.06
LVEF (for 10 points % decrease)1.491.23–1.81<0.001
Duration of HF (for 24 months increase)1.431.19–1.73<0.001
Beta-blockers (yes vs. no)0.550.35–0.880.01
ICD/CRT (yes vs. no)0.610.34–1.090.09
New LBBB (time-dependent) (yes vs. no)3.181.90–5.31<0.001
  • CI, confidential interval; HF, heart failure; HR, hazard ratio; LVEF, left ventricular ejection fraction; MR, mitral regurgitation; NYHA, New York Heart Association functional stage; SBP, systolic blood pressure.

Figure 3

Landmark analysis examining the time-dependent association between new-onset LBBB and the mortality/urgent HTx rate. The horizontal line indicates HR = 1. LBBB, left bundle branch block.

Similar results on the impact of new-onset LBBB were found in the cohort of patients who did not undergo device implantation during follow-up (for details of case selection, see the Statistical analysis section) in both the univariable Kaplan–Meier log-rank test (total mortality/HTx: 45 vs. 26%, P = 0.01) and multivariable Cox model (HR 3.24, 95% CI: 1.87–5.61, P < 0.001).

Discussion

To our knowledge, this is the first study that provides an extensive evaluation of the influence of LBBB on long-term prognosis in a large, single-centre, well-characterized population of young patients affected by DCM.

Left bundle branch block is common among patients with HF, its prevalence depending on the severity of the HF.8,22 In our study, it was present in approximately one-third of the patients at first presentation.

Left bundle branch block has a negative impact on the cardiac performance. However, the published data exploring its impact on the outcome in patients with HF are not conclusive.810,1315 In a study by Tabrizi et al.,10 that included the patients with symptomatic HF requiring hospitalization, LBBB was not an independent predictor of mortality. Conversely, Baldasseroni et al.8 found that the presence of LBBB increased the risk of 1 year all-cause mortality by 36% in a cohort of patients from the Italian Network CHF Registry.

Brembilla-Perrot et al.13 have analysed this issue more recently in patients with DCM. However, they only included the patients with DCM in NYHA I–II, and reported higher event rates during follow-up in patients with LBBB or right bundle branch block. Grimm et al.15 analysed the predictors of survival of 343 patients with DCM. In this study were excluded patients in NYHA Class IV and LBBB has no impact on the subsequent outcome.

In our study, patients with baseline LBBB had worse long-term survival. However, this conduction disorder was not a marker of adverse prognosis at the multivariable analysis. The duration of the disease and severity of LV dysfunction were found to be the strongest determinants of prognosis. Our results differ from other cited studies8,13,14 possibly because the prognostic importance of risk factors may differ according to the stage of myocardial disease, and to medical treatment of the investigated populations.3,23,24 Furthermore, our study cohort has a high prevalence beta-blocker treatment, which was shown to influence the prognostic value of variables such as LBBB used in risk stratification of patients with HF.25

New-onset left bundle branch block and prognosis

The incidence of LBBB is about 0.6% in the general population, and it increases dramatically after 50 years of age.26 To date, there are no data regarding the incidence and predictive value of new-onset LBBB in young patients with HF affected by DCM.

The conduction abnormalities are a part of the natural history of DCM.27 Myocardial fibrosis, which is frequently found in DCM,28,29 has been shown to correlate with complete LBBB.30 During a long-term follow-up, about 8% of our patients with DCM developed LBBB; nearly half of them during the 1 year of follow-up. We did not find any correlation between the previous history of myocarditis and LBBB development (P = 0.25). Also, the familial form of DCM was not significantly associated with the development of LBBB (P = 0.81).

Clark et al.9 analysed a population of outpatients with a chronic stable HF, and found an incidence of new-onset LBBB of 10.4% in the 1 year of follow-up. The incidence of new-onset LBBB in our study was lower than in Clark's experience,9 possibly due to the younger age of our patients.

According to our analysis, the occurrence of new-onset LBBB is associated with a worse outcome, especially when it appears after more than 1 year of medical treatment with ACE inhibitors and beta-blockers. The same was also observed in the cohort of patients who did not undergo CRT–ICD implantation during the follow-up. Similarly, Clark et al.9 observed that LBBB which develops within the 1 year of follow-up has no impact on the outcome at multivariable analysis.

It is not clear why LBBB should be associated with a worse prognosis only when it occurs more than 1 year from diagnosis. New-onset LBBB in association with the pre-existent cardiac remodelling due to DCM may interact in a vicious circle leading to LV wall stress and dilatation, and a progressive worsening of LV function. This deterioration may be more pronounced when compared with the patients with pre-existing LBBB.31 In addition, an early onset of LBBB (within the first year) may identify patients with more advanced myocardial disease (as suggested by longer duration of HF), and a poorer haemodynamic profile (as suggested by the higher rates of diuretic treatment reduced renal function, and moderate-to-severe mitral insufficiency) in which pharmacological optimization has not yet been reached. Therefore, in this group, the appearance of an intraventricular delay may have little prognostic effect. Conversely, after a clinical stabilization and optimization of medical therapy, the development of LBBB could represent an early marker of progression of the primitive myocardial disease.

Cardiac resynchronization therapy has a favourable effect on ventricular remodelling in patients with ventricular conduction abnormalities,3234 and decreases adverse events57 particularly when associated with an ICD.35 Current guidelines36 indeed emphasize the prognostic role of LBBB and extend indications for CRT in patients with mild-to-moderate HF.37,38

Our study gives an important contribution in this issue in the setting of young patients with HF affected by DCM. Data from our study suggests that the development of LBBB in patients who are already on treatment with ACE inhibitors and beta-blockers are associated with a worse prognosis. A tight clinical evaluation and optimization of the medical treatment (dosage of drugs) could have a positive impact on prognosis. Moreover, the consideration that patients with new LBBB experienced a higher SD/MVAs as well as a higher mortality for pump failure maybe because of an early implantation of a CRT-D device (3–9 months after optimal medical therapy39) could maybe improve their prognosis.

Study limitations

Although this is a retrospective study, it assessed an extensive cohort of consecutive patients with DCM enroled in a single-centre Registry with rigorous inclusion criteria, on optimal medical treatment and with a long-term follow-up. Electrocardiogram was only performed at follow-up visits, thus we are unable to determine the exact date when the new-onset LBBB occurred. The conduction disturbances are associated with more advanced stages of HF. The development of LBBB and its specific role in long-term prognosis may thus be difficult to examine. In our study, new-onset LBBB occurring after more than 1 year of follow-up was independently associated with a poorer outcome when considered as a time-dependent variable, and after adjustment for all other clinical prognostic baseline variables. The low prevalence of device therapy in our population was due to the inclusion of patients evaluated more than 10 years ago when CRT/ICD were not evidence-based treatments yet. We only observed a trend towards a protective effect of ICD/CRT treatments on survival, probably because of a lack of power due to the small proportion of treated patients.

Cardiac resynchronization therapy/ICDs were implanted more frequently in patients with new-onset LBBB. The implantable devices record both the symptomatic and asymptomatic arrhythmias. Thus, we cannot exclude the fact that the higher rate of MVA we observed in patients with new-onset LBBB is partly due to the higher sensitivity in their detection by implantable devices.

In conclusion, LBBB is frequent among the patients with DCM, but its presence at the baseline is not an independent marker of poorer survival. The development of LBBB in patients who are already on treatment with ACE inhibitors and beta-blockers is independently associated with an adverse outcome. This occurrence may warrant a more aggressive treatment possibly including early device therapy.

Acknowledgements

The authors acknowledge Fondazione Generali and Associazione Amici del Cuore for continuous and fundamental support to the Registry and Scientific activities.

Conflict of interest: none declared.

References

View Abstract