The benefits of cardiac resynchronisation therapy (CRT) in patients with heart failure (HF) and sinus rhythm (SR) are well established. In the CArdiac Resynchronisation Heart Failure (CARE-HF) study, CRT was associated with 40% relative reduction in all-cause mortality.1 This and other studies have also shown that in patients in SR, CRT leads to an improvement in symptoms, exercise capacity and quality of life, as well as reverse left ventricular (LV) remodelling.2^–6

Heart failure is commonly associated with atrial fibrillation (AF), which is present in 10–25% of patients with HF in New York Heart Association (NYHA) class II–III and in up to 50% of patients in NYHA class IV.7 Heart failure and AF are mechanistically linked through a variety of mechanisms, including tachycardia-induced cardiomyopathy,8 a rate-independent reduction in cardiac output9 and loss of atrioventricular synchrony.10

In patients with AF, loss of the atrioventricular component of the treatment occurs and, therefore, CRT works only by correcting interventricular and intraventricular dyssynchrony. The effectiveness of CRT in AF is further undermined by higher intrinsic ventricular rate responses, which lead to incomplete delivery of CRT. All these factors would be expected to negate the effects of CRT. Notwithstanding, some studies have shown that in patients in AF, CRT improves symptoms,11^–15 whereas others have suggested that CRT is only effective after atrioventricular junction ablation.16 17

Our study was undertaken because the recent National Institute of Clinical Excellence (NICE) Health Technology Appraisal18 19 specifically excludes patients in rhythms other than sinus. We have compared the effects of CRT on mortality and morbidity in consecutive non-ablated patients in AF with results for patients in SR. The effects on symptoms and LV remodelling were also assessed.

PATIENTS AND METHODS

Patients

Consecutive patients who underwent successful biventricular pacemaker implantation were included in this study. Inclusion criteria were HF in NYHA class III or IV; impaired LV function (LV ejection fraction (LVEF) ⩽35%); a QRS complex ⩾120 ms; optimal tolerated drug treatment. Exclusion criteria included contraindications to cardiac pacing; myocardial infarction or acute coronary syndrome within the previous 3 months; presence of, or indications for, an implantable cardioverter-defibrillator; severe structural valvular heart disease; presence of comorbidities likely to threaten survival for 12 months. All participants had undergone coronary angiography. The clinical diagnosis of HF was made on the basis of documented evidence of systolic dysfunction on echocardiography. The diagnosis of ischaemic cardiomyopathy was made if systolic dysfunction was associated with a history of myocardial infarction20 and if there was angiographically documented coronary heart disease (>50% stenosis in ⩾1 coronary arteries). All patients gave written informed consent and the study was approved by the North Birmingham Ethics Committee.

Study design

Patients referred for CRT underwent pacemaker implantation during an elective admission or after stabilisation during the course of an unplanned admission for acute decompensated HF. Patients underwent a clinical assessment and echocardiography the day before implantation and at 1, 3 and every 6 months after pacemaker implantation. Clinical follow-up data relate to the last available follow-up before death or the end of the study period.

Clinical assessment

Clinical assessment included evaluation of NYHA functional class, quality of life assessment using the Minnesota Living with Heart Failure questionnaire21 and a 6-minute hall-walk test.22 AF was considered paroxysmal if episodes had terminated spontaneously within 7 days. AF that could not be successfully terminated by cardioversion, and longstanding (>1 year) AF, where cardioversion was not indicated or had not been attempted, was termed permanent.23

Echocardiography

Two-dimensional echocardiography was performed using a Vivid System 5 scanner with EchoPAC (General Electric Worldwide, Slough, UK). Standard left parasternal long-axis and short-axis, and apical, four-, five- and two-chamber views were obtained. Digital images were stored for offline analysis. LV volumes were estimated through the Simpson equation by planimetry of apical four-chamber views. The frame at the beginning of the QRS complex was used to estimate LV end-diastolic volume (LVEDV), while the frame with the smallest ventricular area just before mitral valve opening was used to estimate LV end-systolic volume (LVESV).

Device therapy

Patients underwent transvenous biventricular pacemaker implantation, which was undertaken using standard techniques under local anaesthesia. Patients who were in SR after implantation underwent transmitral, Doppler-directed optimisation of atrioventricular delay24 before discharge and at every scheduled visit thereafter. Backup atrial pacing was set at 60 bpm, and the pacing mode was set to DDDR with an interventricular delay of 4 ms. For patients in AF, right atrial leads were only implanted if there was evidence of SR in the 2 years before implantation. All patients with AF were programmed to ventricular triggered mode. All patients in the study were programmed to a minimum rate of 80 bpm and a maximum of 130 bpm. Generators used included the Medtronic InSync models 8040 and 8042, St Jude Frontier, Vitatron CRT 8000, Biotronik Stratos, and the Guidant Contak TR model 1241 and Renewal TR2. Patients were entered into the study only after a successful implantation and were followed up in a dedicated CRT clinic.

End points

The primary end point considered was the composite of cardiovascular death or an unplanned hospitalisation for a major cardiovascular event (MCE), which included cardiac transplantation. Admissions to hospital for worsening HF, myocardial infarction, unstable angina, arrhythmia, stroke, pulmonary embolism or upgrading to an implantable cardioverter-defibrillator were included in this end point. The first event was included in the analysis. The secondary end point considered was the composite of cardiovascular death or unplanned admission to hospital with worsening HF. Additional end points were cardiovascular mortality and mortality from all causes. Mortality data were collected through medical records, and where appropriate, from interviews with patients’ carers. Information about clinical outcome was collected by an investigator who was unaware of the clinical and echocardiographic variables. Sudden cardiac death was defined as “a natural, unexpected death due to cardiac causes, heralded by an abrupt loss of consciousness within 1 hour of the onset of acute symptoms”.25

Statistical analysis

Continuous variables are expressed as mean (SD). Normality was tested using the Shapiro–Wilk test (the W-statistic). Comparisons between normally distributed continuous variables were made using analysis of variance with the Scheffe F procedure for multiple comparisons. Serially measured variables were analysed using repeated measures analysis of variance. Categorical variables were analysed using χ² tests.

The effect of AF on the various end points was explored using Kaplan–Meier survival curves. Differences in survival curves between the groups were assessed using the log-rank (Mantel–Cox) test. Statistical analyses were performed using Statview (Cary, NC, USA) and SPSS 13.0 (Chicago, Illinois, USA). A two-tailed p value of <0.05 was considered significant.

RESULTS

Table 1 shows that patients with AF were well matched to those in SR. β-Blockers were used to the same extent in both groups. The mean follow-up for the clinical end points was 691 days (range 30–2470). In patients who died, the clinical and echocardiographic data at follow-up pertain to the latest data available at follow-up, at 575 days (range 38–2050) after implantation.

Clinical variables

Table 2 shows that NYHA class, 6-minute walking distance and quality of life scores were comparable between the AF and SR groups. Up to the last available follow-up, the AF and SR groups derived similar improvements in mean NYHA class (−1.3 vs −1.2), mean 6-minute walking distance (92.3 vs 78.4 m) and mean quality of life scores (−25.2 vs –18.7) (all p<0.001).

Echocardiography

Table 2 shows that evidence of reverse LV remodelling emerged in both the AF and SR groups. There were reductions in LVESV (−25.9 vs –34.5 ml, both p<0.001) and LVEDV (−20.2 ml (p = 0.001) vs 26.2 ml (p<0.001)) and improvements in LVEF (4.69% vs 7.86%, both p<0.001).

End points

Table 3 and fig 1 show that no differences between the AF and SR groups emerged with respect to the composite end points of cardiovascular death or unplanned hospitalisation for MCEs or the composite end point of cardiovascular death or hospitalisation for HF. Moreover, no differences emerged with respect to cardiovascular mortality or mortality from all causes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Kaplan–Meier estimates of the time to the various clinical end points in patients in atrial fibrillation (AF) and patients in sinus rhythm (SR). HF, heart failure; MCE, major cardiovascular event. No significant group differences emerged with respect to any of the end points.

DISCUSSION

We have shown that patients with HF and AF undergoing CRT derive similar prognostic benefits to those of patients in SR. Over a median follow-up period of 23 months, we have seen no difference between patients in AF and patients in SR with respect to the composite end points of death or unplanned hospitalisation for MCE, death from any cause or unplanned hospitalisation for HF, death from any cause or cardiovascular death. Improvements in NYHA class, 6-minute walking distance and quality of life were also similar. In addition, there was evidence of reverse LV remodelling in both groups.

The development of AF in patients with HF heralds a poor prognosis,26 27 with the mortality risk doubling after its onset.28 It is therefore surprising to find that after CRT, mortality and morbidity end points in patients with AF are similar to those in SR. Our findings are in keeping with those of a post hoc analysis in the CARE-HF trial showing that CRT reduced the risk all predefined mortality end points, regardless of whether AF developed over the follow-up period.15 As AF is associated with a higher mortality, the finding of similar mortality in the AF group and in patients with SR raises the possibility that CRT may confer a proportionally greater prognostic benefit in patients with AF.

Our demonstration of a symptomatic benefit from CRT in patients with AF is in keeping with the findings of the MUltisite STimulation In Cardiomyopathies (MUSTIC) study, in which 42 patients in SR and 33 patients in AF undergoing CRT derived similar improvements in NYHA class, 6-minute walking distance, peak oxygen uptake and quality of life at 12 months.11 Comparing 30 patients in SR and 30 patients in AF, Molhoek et al also found significant improvements in NYHA class, 6-minute walking distance and quality of life scores after 2 years in both AF and SR groups, although the responder rate was lower in the AF group.12 Recently, in a study of 74 patients in AF (20 paroxysmal and 54 permanent) undergoing CRT, Kies et al showed significant improvements in NYHA class, quality of life scores and 6-minute walking distance at 6 months.13 Similar findings were recently reported at 12 months after implantation in a study of 96 patients with permanent AF by Delnoy et al.29 In the only randomised trial to include patients with AF, 39 patients with AF were shown to have a similarly improved walking distance as patients in SR.14 In the CARE-HF trial, 124/813 patients (15%) developed AF over a mean follow-up of 29.4 months.15 In these patients, CRT significantly improved symptoms, regardless of whether AF developed over the follow-up period.

The reverse remodelling effects of CRT in patients in SR have been shown by numerous studies.2 11 30 With respect to patients in AF, Kies et al showed that CRT led to a reduction in LV end-diastolic and end-systolic diameters.13 In another study of 16 patients with AF treated with the “ablate-and-CRT” approach, upgrading to a biventricular pacemaker was associated with a significant reduction in LV end-systolic diameter and an increase in fractional shortening at 20 months.31 In the present study, we have shown that, compared with patients in SR, patients in AF exhibit comparable reductions in LV volumes. Moreover, these changes paralleled an improvement in LVEF, which was also similar. This finding is consistent with the CARE-HF study, in which CRT led to an increase in LVEF, regardless of whether AF developed during the follow-up period.15

The most salient finding from this study is that, after CRT, patients with AF and without prior atrioventricular junction ablation derived similar benefits to patients in SR. Although the percentage biventricular pacing achieved at follow-up in patients with AF was less than in patients with SR, it was, nevertheless, high, at 87%. Some observational studies have suggested that CRT is only beneficial after atrioventricular junction ablation, so as to achieve a high percentage biventricular pacing.16 17 Gasparini et al explored the effects of CRT in 48 patients with permanent AF taking rate-slowing drugs, 114 patients with permanent AF who had undergone atrioventricular junction ablation and 511 patients in SR.16 After a mean follow-up of 25 months, only patients in the AF group who underwent atrioventricular junction ablation showed a significant improvement in exercise tolerance and reverse LV remodelling. The absence of a benefit in non-ablated patients with AF is in contrast to the findings of two other studies,11 14 as well as the present study, all of which show improvements in exercise tolerance as well an increase in LVEF in patients in AF undergoing CRT. Importantly, the study of Gasparini et al was not randomised, and did not quantify mortality or morbidity. Moreover, it is possible that patient selection as well as continuation of rate-slowing treatment in the “ablate-and-CRT” arm may have influenced the outcome.32 An important demonstration of studies using the “ablate-and-CRT” approach is that the interventricular and intraventricular synchronisation provided by CRT is sufficient to improve cardiac performance in patients with AF.16 17 31 33 Only a randomised study of CRT in patients with AF with or without prior atrioventricular junction ablation is likely to resolve these issues.

Limitations of the study

This study is observational. Our findings of a comparable prognosis and improvements in symptoms as well as echocardiographic variables, though consistent, require confirmation in a randomised study. As all patients were programmed using the ventricular triggered mode, we have not been able to ascertain whether some of the effects seen in this study were attributable to the use of this pacing mode. This would require inclusion of a group who had not been programmed using this mode. A systematic, prospective assessment of AF burden or mode-switching to VVIR mode was not undertaken. Arguably, the occurrence of AF in patients with SR at follow-up might have influenced the outcome in this group.

CONCLUSIONS

We have shown that CRT leads to similar prognostic benefits in patients in AF or SR. The finding of a symptomatic benefit is consistent with the finding that, as for patients in SR, CRT also leads to reverse LV remodelling in patients with AF. Most importantly, mortality and morbidity in patients with HF and AF who undergo CRT were, unexpectedly, comparable with results for patients in SR. Whether or not atrioventricular junction ablation should be mandatory for patients in AF undergoing CRT requires further evaluation in a randomised study.