Skip Navigation

European Journal of Heart Failure 2007 9(6-7):730-739; doi:10.1016/j.ejheart.2007.01.013
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (4)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Cleland, J. G.F.
Right arrow Articles by Grind, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cleland, J. G.F.
Right arrow Articles by Grind, M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 2007 European Society of Cardiology

Prevalence of markers of heart failure in patients with atrial fibrillation and the effects of ximelagatran compared to warfarin on the incidence of morbid and fatal events: A report from the SPORTIF III and V trials

John G.F. Clelanda,*, Rhiddian Sheltona, Nikolay Nikitina, Sarah Forda, Lars Frisonb and Margaretha Grindb

a Department of Cardiology, Kingston-upon-Hull, UK
b AstraZeneca, Mölndal, Sweden

* Corresponding author. Department of Cardiology, University of Hull, Castle Hill Hospital, Kingston-upon-Hull, UK. Tel: +44 1482 624 087; fax: +44 1482 624 085. E-mail address: j.g.cleland{at}hull.ac.uk


   Abstract
Top
 Abstract
1. Introduction
 2. Methods
 3. Results
 4. Discussion
 Acknowledgment
 References
 
Background: Patients with atrial fibrillation (AF) who also have heart failure have a worse outcome but the diagnosis of heart failure is often missed.

Aim: To compare the effects of warfarin and ximelagatran on morbidity and mortality in patients with AF with and without markers of heart failure.

Methods and results: Data for 7329 patients from two randomised controlled trials were merged. Treatment with loop diuretics or the presence of left ventricular dysfunction, were used as markers of possible heart failure. The 3555 (49%) patients with markers of heart failure had higher composite event rates on warfarin (10.81% per year [py] [95% CI 9.59 to 12.13]) compared to the 3774 (51%) patients without markers of heart failure (4.18% py [95% CI 3.44 to 5.01]). The composite event rate was lower on ximelagatran overall (6.18% py [95% CI 5.51 to 6.89] versus 7.34% py [95% CI 6.63 to 8.10] on warfarin; P=0.0219 for the difference) with similar effects in each trial and in patients with and without markers of heart failure, mainly due to fewer heart-failure events (hazard ratio 0.69 [95% CI 0.54 to 0.87]; P<0.001).

Conclusions: Patients with markers of heart failure, even if the diagnosis is not well established, are at increased risk of thromboembolic events and might be targeted for more effective antithrombotic therapy. This might include patients in sinus rhythm as well as AF.

Key Words: Anticoagulants • Heart failure • Atrial fibrillation • Ximelagatran • Warfarin

Received July 21, 2006; Revised December 5, 2006; Accepted January 31, 2007


   1. Introduction
Top
 Abstract
 1. Introduction
2. Methods
 3. Results
 4. Discussion
 Acknowledgment
 References
 
Many patients with atrial fibrillation (AF) have major structural heart disease [1]. Many other patients with AF are prescribed loop diuretics, presumably for symptoms or signs of fluid congestion, without investigation for objective evidence of structural heart disease or a diagnosis of heart failure being made. Many of these patients, if investigated, will have preserved left ventricular systolic function and left atrial dilatation but do not fulfil conventional Doppler criteria for diastolic heart failure [2,3]. Conversely, about 30% of patients with heart failure will be in AF [1] and, for some, AF will be the only obvious cause [4].

Mortality and morbidity rates are higher in patients who have AF [5] when it is complicated by heart failure, and the onset of AF confers a worse prognosis amongst patients with heart failure due to left ventricular systolic dysfunction (LVSD), even if the patients are treated with contemporary pharmacologic therapy including angiotensin-converting enzyme (ACE) inhibitors, beta-blockers, digoxin and anticoagulants [6]. It is unclear whether patients treated with loop diuretics, but who have not been shown to have underlying LVSD, have a different risk of events from patients in whom LVSD has been demonstrated.

There are risks inherent in anticoagulant therapy and therefore the risks of the condition being treated must be substantial in order to justify therapy. It is amongst patients with AF complicated by heart failure that the benefits of warfarin therapy are most clear-cut but good anticoagulant control may be problematic in such patients for a number of reasons [7]. Varying degrees of hepatic congestion and drug interactions, including amiodarone for arrhythmias, and antibiotics, which are frequently prescribed due to the greater rate of chest and other infections, make anticoagulant control more difficult. On the other hand, anaemia is common in patients with heart failure [8]. This is usually a normochromic normocytic anaemia, which is common in several chronic disease states but patients with heart failure are at increased risk of gastrointestinal bleeding that may exacerbate anaemia by causing iron deficiency and will increase concern about the use of anticoagulants. Thus, the potential risks as well as benefits of therapy are increased in patients with heart failure and AF.

Ximelagatran, an orally active direct thrombin inhibitor, was the first of a new class of anticoagulant that did not require regular monitoring of coagulation, had no food interactions and a low potential for drug interactions. Two large randomised controlled trials, SPORTIF III [9,10] and SPORTIF V, [10-12] provided information on a large number of patients with non-valvular AF followed for a substantial period and suggested that the risk of death and major vascular events was similar on ximelagatran and warfarin. This analysis compares the efficacy and safety of ximelagatran with warfarin in patients with non-valvular AF, with and without evidence of heart failure. Although ximelagatran has been withdrawn because of fears about hepatic safety, other agents of this class are now being developed and therefore this analysis is still of considerable interest.


   2. Methods
Top
 Abstract
 1. Introduction
 2. Methods
3. Results
 4. Discussion
 Acknowledgment
 References
 
The main objective of this report is to compare the effects of ximelagatran and warfarin on important clinical outcomes and adverse events in patients with non-valvular AF, with and without markers of heart failure. A secondary objective was to investigate the effect of markers of heart failure on outcome in a contemporary population of patients with AF in a clinical trial. These analyses were specified only after publication of the principal outcome for each trial.

The patient characteristics, and primary and secondary outcomes of SPORTIF III and SPORTIF V have been published [9-11]. The key inclusion and exclusion criteria for each of these prospective randomised controlled trials were identical. There were two main differences between the studies. In the SPORTIF III trial, conducted in Europe, Asia and Australasia, patients and investigators were not blinded to treatment allocation, which was therefore an appropriate test for clinical practice of two different anticoagulant strategies. In the SPORTIF V trial, conducted in the United States and Canada, patients and investigators were blinded to treatment allocation, which was therefore a robust test of the effects of two pharmacologic agents.

The main inclusion criteria were age ≥18 years, persistent or paroxysmal non-valvular AF (NVAF) and at least one additional risk factor for stroke including: previous stroke, transient ischaemic attack (TIA) or systemic embolism, hypertension, left ventricular dysfunction (LVD, defined as an ejection fraction <40% or symptomatic systolic or diastolic heart failure in the investigators opinion — no specific criteria were set), aged ≥75 years or aged ≥65 years with known coronary artery disease or diabetes mellitus.

The main exclusion criteria were: a stroke, systemic embolic event or acute coronary syndrome within 30 days or TIA within 3 days before inclusion; an increased risk of bleeding or clot formation, such as a mechanical heart valve or intracardiac thrombus; AF secondary to reversible disorders (e.g., thyrotoxicosis); planned cardioversion; planned major surgery; antiplatelet therapy (except aspirin ≤100 mg/day) or regular use of nonsteroidal anti-inflammatory drugs; hypertension >180/100 mm Hg; a calculated creatinine clearance <30 mL/min using the Cockroft-Gault formula; or persistently elevated liver enzymes ≥2 times the upper limit of normal. Other exclusion criteria have been published [9-11].

The protocol was approved by the appropriate ethics committees for each institution. Written informed consent was obtained from all patients.

Investigators were asked if the patient had LVD, as defined above, and to provide information on baseline medication, including loop diuretics. These data were verified from source documents by Clinical Research Associates during the course of the study. However, heart failure is a frequently overlooked condition that is commonly treated with loop diuretics without adequate investigation leading to diagnosis [1,13]. As patients with major renal dysfunction were excluded from this study, it is likely that loop diuretics were being used predominantly to control symptoms and/or signs of heart failure. Therefore, while use of loop diuretics is not diagnostic of heart failure, their use does indicate a population with a higher prevalence of this condition. The presence of LV dysfunction in conjunction with the use of loop diuretics is even more suggestive of heart failure. Using these criteria we divided the trial populations into the following four groups of patients.

  1. Those not known to have LVD and not taking loop diuretics. The likelihood of moderate or severe heart failure in these patients is low.
  2. Those with LVD although not necessarily heart failure and not taking loop diuretics.
  3. Those taking loop diuretics, many of whom will be taking this treatment for the control of symptoms and signs of heart failure but in whom a diagnosis of heart failure has been missed.
  4. Those with LVD also taking loop diuretics, most of whom will have heart failure.

The primary pre-specified outcome of each trial in each original analysis plan was the composite of stroke and systemic embolic event (SEE) using a time to first event analysis and by intention-to-treat. For the current analysis, several new outcomes were specified. The nominal primary outcome of this post-hoc analysis was the composite of all-cause mortality, stroke, myocardial infarction, SEE and heart-failure related hospitalisation verified by source documents. Other outcomes were a) all-cause mortality or heart-failure related hospitalisation or b) all-cause mortality or any report of a serious adverse event of worsening heart failure. These analyses were done in the overall population, in patients without reported LVD or loop diuretic use, in whom a low prevalence of heart failure would be expected, and in patients either with reported LVD or who were taking loop diuretics; in whom a high prevalence of heart failure would be expected.

A serious adverse event report for heart failure included worsening symptoms, signs, test results or hospitalisation related to heart failure. Hospitalisation with heart failure was defined as an adverse event report indicating heart failure as the reason for hospitalisation or prolongation of hospital stay. All reports were validated from source documents during the study. Patients who withdrew from therapy or who reached a primary end point were subsequently followed only for death, stroke and SEE. Accordingly, there is a small amount of missing data for some of the new secondary outcome measures and for the rate of heart failure events after the primary end point, which should be taken into account when interpreting results.

Rates of bleeding (major only or minor and major combined) and of abnormalities of liver enzyme levels, a potential side effect of ximelagatran, were also assessed.

The patients were scheduled to attend follow-up clinics at 1, 4, 6 and 8 weeks after randomisation and then monthly until 6 months, every 2 months thereafter until 1 year and then quarterly. Patients also attended for checks as required for clinical purposes and for anticoagulant monitoring. A central laboratory was used for all laboratory analyses except for International Normalized Ratio (INR) values in SPORTIF III, where local labs were used. In patients randomised to warfarin in SPORTIF III and in both groups in SPORTIF V, the INR was checked at least every 30 days in order to achieve a value between 2 and 3.

2.1. Statistical analysis
The primary analysis compared the effects of warfarin and ximelagatran on the composite end point for this report of death, stroke (ischaemic or haemorrhagic), systemic embolism, myocardial infarction or heart-failure related hospitalisation. The scheduled duration in the studies differed between patients according to differences in time of enrolment, since the period for study closure was the same for all patients. Annualized event rates were calculated, and formed the basis for analyses, assuming constant event rates over time. All randomised patients were included in the intention-to-treat population. All patients were followed for occurrences of stroke, systemic embolism and mortality until study closure, while other end-points were recorded only while patients were receiving study drugs. Composite end-points confined to stroke, systemic emboli, and death were analyzed according to intention-to-treat. Assessments of all other end-points and other composite end-points used on-treatment approach unless otherwise stated. Time in on-treatment (OT) analysis was censored following interruption of therapy for ≥30 days (or ≥60 days in the case of cardioversion) consecutively, or ≥60 days cumulatively. The study design required that patients stop study drug if they had a non-fatal stroke or systemic embolic event. Fisher's exact test was used to compare groups with a two-tailed approach and 95% confidence intervals. Hazard ratios between the treatments were obtained through Cox-regression modelling.


   3. Results
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
4. Discussion
 Acknowledgment
 References
 
Altogether, 7329 patients were included in this analysis: 3407 from SPORTIF III and 3922 from SPORTIF V. The characteristics of patients in these two studies were similar and have been published elsewhere [9-11]. The mean age of the patients was 71 years and 31% were women (Table 1). Patients had a similar number of clinic contacts, mainly protocol driven, and therefore ascertainment bias for non-fatal outcomes appears unlikely. Patients had a mean of 13 visits, regardless of treatment assignment or the study in which they were enrolled.


View this table:
[in this window]
[in a new window]

  		Table 1 Patient characteristics (number and % of patients) at trial entry according to use of loop diuretics and presence of LVD in the intention-to-treat population

 
At baseline, LVD was reported in 2681 patients (37%), 1825 of whom (68%) were taking loop diuretics. Eight hundred and seventy-four patients (12%) were taking loop diuretics but were not known to, or known not to, have LVD (Fig. 1). Almost half of the patients (49%) were taking loop diuretics or had LVD or both. Patient characteristics are shown in Table 1. Clinically relevant differences are highlighted in bold and all represent highly significant differences (P<0.001). Patients with LVD were more likely to have a history of coronary disease but less likely to have a history of hypertension, had lower systolic blood pressure (SBP) and were more likely to be treated with loop diuretics, ACE inhibitors, beta-blockers and digoxin compared to patients without LVD and not taking loop diuretics. Patients taking loop diuretics without known LVD were older, more often women and were more likely to have diabetes, a history of hypertension and coronary disease and to take an ACE inhibitor and digoxin. Few patients were receiving thiazides, especially if they were taking a loop diuretic.


Figure 01
View larger version (10K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 Venn diagram showing how patients were classified into subgroups and their proportions. Patients were classified according to LVD reported by the investigator and use of loop diuretics.

 
The rate of the original primary end point, stroke or SEE, was similar in all four groups (Table 2). Compared to patients without known LVD and not taking loop diuretics, the revised primary composite outcome and each of its components was considerably higher in patients with known LVD taking loop diuretics and increased to a lesser extent in patients with only one of these markers (Table 2; Fig. 2).


View this table:
[in this window]
[in a new window]

  		Table 2 Number of patients with an event, patient-years exposure and rate of events per year of follow-up classified by markers for heart failure

 


Figure 02
View larger version (13K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 Kaplan-Meier curves (on the same Y-axis scale) showing time to the revised primary end point for: a) patients with both LVD and using loop diuretics; b) patients using loop diuretics only; c) patients with LVD only; and d) patients with no LVD and not taking loop diuretics. Data for patients assigned to warfarin or ximelagatran have been combined. For event rates see Table 2.

 
The mean duration of follow-up was 18.5 months for the intention-to-treat analysis and 16.4 months for the on-treatment analysis. In the overall population (Table 3), patients assigned to warfarin and ximelagatran had a similar incidence of stroke or systemic embolism, the original primary end point analyzed on an intention-to-treat basis (Fig. 3). However, the incidence of the post-hoc primary end point was lower with ximelagatran compared to warfarin (warfarin 7.34% per annum [95% CI 6.63 to 8.10] versus ximelagatran 6.18% per annum [95% CI 5.51 to 6.89) P=0.0219], an effect that was due mainly to a reduction in heart-failure related hospitalisation (Fig. 3). The incidence of death or heart-failure related hospitalisation was also lower on ximelagatran compared to warfarin (warfarin 5.49% per annum [95% CI 4.88 to 6.16] versus ximelagatran 4.35% per annum [95% CI 3.79 to 4.96; P=0.009]). The observed effects were similar in SPORTIF III (unblinded) and V (blinded) (Fig. 4). The hazard ratio for heart-failure related hospitalisation on ximelagatran versus warfarin was 0.69 (95% CI 0.54 to 0.87) overall, 0.56 (95% CI 0.37 to 0.83) in SPORTIF III and 0.79 (95% CI 0.58 to 1.06) in SPORTIF V. The hazard ratio for death or heart-failure related hospitalisation overall was 0.79 (95% CI 0.66 to 0.95), 0.76 (95% CI 0.57 to 1.00) in SPORTIF III and 0.82 (95% CI 0.65 to 1.04) in SPORTIF V.


View this table:
[in this window]
[in a new window]

  		Table 3 Patient characteristics at trial entry according to randomised group and the presence or absence of markers of cardiac dysfunction in the intention-to-treat population

 


Figure 03
View larger version (12K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3 Kaplan-Meier curves showing time to first event for: A) the original primary end point (HR=0.98 [CI 0.74 to 1.31]); B) the revised primary end point (HR=0.84 [CI 0.72 to 0.98]); C) death or an adverse heart-failure event (HR=0.79 [CI 0.68 to 0.91]); and D) death or hospitalisation for heart failure (HR 0.79 [95% CI 0.66 to 0.95]), in patients taking warfarin or ximelagatran in the SPORTIF III and V populations combined.

 


Figure 04
View larger version (6K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4 The revised primary end point in SPORTIF III (left-hand panel) (overall HR 0.75 [95% CI 0.60 to 0.95]) and V (right-hand panel) (overall HR 0.92 [95% CI 0.75 to 1.13]).

 
The observed reduction in heart-failure related events with ximelagatran was proportionally similar in patients with or without markers for heart failure at baseline (Table 4; Fig. 5). However, patients who did not have LVD and who were not taking loop diuretics had fewer events (Table 4).


View this table:
[in this window]
[in a new window]

  		Table 4 Number of patients with an event and rate of events per year of follow-up in patients with or without evidence of heart failure

 


Figure 05
View larger version (9K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5 Kaplan-Meier curves showing time to first event for: A) the revised primary end point in patients without either LVD or loop diuretics (CHF–ve) versus those with either/or LVD and loop diuretics (CHF+ve); and B) death or an adverse heart-failure event in patients without either LVD or loop diuretics versus those with either/or LVD and loop diuretics in patients taking warfarin or ximelagatran. See Table 4 for hazard ratios and event rates.

 
The following variables were entered into a Cox-regression model to identify independent predictors of heart-failure related hospitalisation: study drug, study, haemoglobin at baseline, sex, age, weight, body mass index, ethnic origin, history of hypertension, previous stroke/TIA, previous SEE, coronary artery disease (CAD), diabetes, aspirin use at entry, smoking, alcohol use, paroxysmal AF, diastolic and SBP, heart rate, and markers of heart failure.

The most powerful independent predictor of heart-failure related hospitalisation was the presence of markers of heart failure at baseline (hazard ratio 55.68; P<0.0001). Use of ximelagatran was also a protective predictor (hazard ratio 0.719; P=0.0033). Lack of use of alcohol, diabetes, lower haemoglobin, greater age and history of CAD were additional weaker independent predictors.

Cox-regression analyses for ACE inhibitors, beta-blockers, aldosterone antagonists and digoxin for the composite outcome of stroke or SEE, major bleeding and the revised primary end point showed no significant interaction with assigned treatment (lowest P-value: 0.16).

Patients with markers of heart failure were at greater risk of any or major bleeding episodes (both P<0.0001) and there were fewer bleeds overall with ximelagatran (P<0.0001). However, no difference in mean haemoglobin values was observed during the course of the study (Fig. 6) amongst patients assigned to ximelagatran or warfarin. Anaemia, defined by World Health Organization criteria, occurred at some time during follow-up in 33.6% of patients on warfarin and 32.2% of patients on ximelagatran. Patients who subsequently died or were admitted with heart failure had lower baseline haemoglobin which fell further during follow-up but again no differences between treatment groups could be observed.


Figure 06
View larger version (9K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 6 Changes in mean haemoglobin over time, in patients randomised to ximelagatran or warfarin who did or did not develop the composite end point of death or hospitalisation for heart failure.

 
Patients taking aspirin were at increased risk of bleeding. In patients taking aspirin, 5.03% per year (py) experienced a major bleed if they were taking warfarin and 2.80% if taking ximelagatran. Amongst patients not taking aspirin, 1.84% py and 1.64% py, respectively, experienced a major bleed. The risk of major bleeds was slightly greater in patients who had markers of heart failure whether taking aspirin (3.52% per annum) or not (1.96% per annum) on ximelagatran. Respective annual rates on warfarin were 5.19% and 2.25%. There was no statistically significant interaction between treatment assignment and increased risk of bleeding with aspirin and no significant interaction between aspirin use and treatment assignment for the primary outcome (P=0.98), or for the composite of death or hospitalisation for heart failure (P=0.30).

An increase in alanine aminotransferase (ALAT) exceeding three times the upper limit of normal (3xULN) was observed in 224/3664 (6.1%) of patients on ximelagatran and 29/3665 (0.8%) of patients on warfarin. Amongst patients with markers of heart failure, an increase in ALAT occurred in 93/1949 (4.8%) on ximelagatran compared to 17/2023 (0.8%) on warfarin and amongst patients without markers, in 131/1715 (7.6%) compared to 12/1642 (0.7%) respectively. The risk of developing ALAT>3xULN for ximelagatran patients was significantly (P=0.0003) lower amongst those with heart failure. This could not be explained by a greater number of deaths or higher baseline ALAT.


   4. Discussion
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
Acknowledgment
 References
 
There is powerful evidence that well-controlled anticoagulation with warfarin reduces the risk of death and stroke in patients with AF [14]. This post-hoc analysis of the SPORTIF III and V trials suggests that, compared to warfarin, use of ximelagatran is associated with similar outcome in terms of death and stroke but a lower risk of heart-failure related events. The observation appeared consistent using different markers for the presence of heart failure, using different criteria for heart-failure events and occurred both in the setting of an open-label and a double-blind trial. Furthermore, the difference could not be accounted for by a difference in reporting, as patients on ximelagatran and warfarin had a similar frequency of contacts with investigators. The mechanism of this effect remains unclear.

Both AF and heart failure are highly prothrombotic states, with evidence of activation of haemostatic systems, vessel wall disease, endothelial dysfunction and reduced blood flow velocity [15,16]. Atherothrombotic events may be a common cause of sudden death, stroke and recurrent myocardial damage leading to worsening heart failure. However, no difference in overt vascular events that might have precipitated worsening heart failure was observed in patients assigned to warfarin or ximelagatran. Patients with heart failure may be prone to ischaemic cardiac events that are not clinically obvious, which could account for increases in serum troponin that are frequently observed during exacerbations of heart failure [17,18] but relevant tests were not done in this study. Evidence of fresh myocardial infarction that was not diagnosed before death is often observed at post-mortem in patients reported to have died of severe heart failure [19,20]. Accordingly it is possible that ximelagatran was more effective at reducing occult occlusive vascular events. However, the lack of effect on death or stroke does not lend support to this hypothesis.

An alternative possibility is that the higher rate of bleeding with warfarin led to an increased rate of anaemia leading to an exacerbation of heart failure. However, no difference in the rate of anaemia or mean haemoglobin level was observed despite detailed analyses. Moreover, despite a marked increase in the risk of major haemorrhage amongst patients taking aspirin, no interaction on clinical outcome between assigned treatment and aspirin was observed.

Warfarin is highly protein bound and it is possible that it interferes subtly with the effects of other cardiovascular medication but no interaction between background treatment and assigned therapy could be identified.

About 50% of patients with AF in the SPORTIF III and V studies had markers suggesting they might have heart failure. This is consistent with the prevalence of heart failure amongst patients with AF reported by some [1,21]. Other studies suggest a lower prevalence of heart failure [14]. The latter reports may reflect the recruitment of younger, fitter patients, especially in trials of cardioversion. However, 11.9% of our patients were receiving a diuretic but had no documented LVD. It is likely that heart failure is under-reported amongst patients with AF, reflecting the difficulty of making an accurate assessment of left ventricular function in this setting due to beat-to-beat variations in ventricular function and the inability to apply conventional Doppler criteria for the diagnosis of diastolic dysfunction in patients with a normal left ventricular ejection fraction [2,3]. Natriuretic peptides are increased in patients with AF whether or not they have other evidence of heart failure, indicating that the diagnostic threshold in this setting is different from patients in sinus rhythm [22]. This analysis shows that patients with AF who are prescribed loop diuretics, whether or not they have LVD, have a high rate of cardiovascular events. Wider recognition of the prescription of diuretics as a possible marker of heart failure and increased risk in patients with AF seems appropriate. The possibility that loop diuretics contribute to adverse outcome should also be considered [23].

Neither warfarin nor aspirin has been shown to be effective in reducing cardiovascular morbidity or mortality in patients with heart failure in sinus rhythm [24,25]. Indeed, aspirin was observed to increase the rate of admissions for heart failure substantially compared to warfarin or no antithrombotic treatment. This may reflect inhibition of endogenous vasodilator prostaglandin synthesis by aspirin and, controversially, could account for the adverse interaction between aspirin and ACE inhibitors [26,27]. The risk of thromboembolic and other cardiovascular events is higher in patients with heart failure in sinus rhythm than in patients with AF who do not have heart failure. Perhaps future studies comparing different anticoagulants should focus on heart failure, regardless of heart rhythm. Scientifically, patients in sinus rhythm should also be randomised to a placebo arm and the feasibility of this approach needs to be assessed [24,25].

Surprisingly, patients with markers for heart failure were less likely to develop abnormal liver enzyme levels on ximelagatran. The difference was highly statistically different and therefore is probably not a chance finding. Inhibition or induction of liver enzymes by heart failure or its treatment might be responsible. This finding should be confirmed by future studies and an explanation sought.

The main limitations of this study are the retrospective nature of the analysis and the fact that some heart-failure events will have been missed due to the study design. Treatment allocation was not blinded in the SPORTIF III trial and although we could identify no bias induced by differences in follow-up, such an effect cannot be discounted. We do not know the proportion of patients with heart failure or receiving diuretics that had a low ejection fraction. On the other hand the study is large and the effect appears consistent in the two trials although less marked in the SPORTIF V trial that was conducted double blind. Ultimately this is a hypothesis-generating exercise to inform future trials and analyses.

In conclusion, patients with AF who do not have LVD and who are not prescribed loop diuretics are at low risk of cardiovascular events whilst receiving warfarin or ximelagatran. Patients taking loop diuretics or who have LVD are at higher risk of events, and in this setting ximelagatran and warfarin appear similarly effective in reducing embolic events. However, ximelagatran is associated with a lower risk of hospitalisation for heart failure. The reasons for this are not clear and do not seem related to differences in the frequency of follow-up or in the development of anaemia, despite a lower risk of haemorrhage on ximelagatran. Future studies investigating the effect of this class of agent should consider focusing on patients with heart failure and including death and heart-failure events as outcomes of interest.


   Acknowledgment
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
 Acknowledgment
References
 
The authors would like to acknowledge the considerable work on the part of the investigators for the conduct of these studies.


   References
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
 Acknowledgment
 References
 

  1. Khand A.U., Rankin A.C., Kaye G.C., Cleland J.G. Systematic review of the management of atrial fibrillation in patients with heart failure. Eur Heart J (2000) 21:614–632.[Abstract/Free Full Text]
  2. Vasan R.S., Levy D. Defining diastolic heart failure: a call for standardized diagnostic criteria. Circulation (2000) 101:2118–2121.[Free Full Text]
  3. European Study Group on Diastolic Heart Failure. How to diagnose diastolic heart failure. Eur Heart J (1998) 19:990–1003.[Free Full Text]
  4. McMurray J., Ostergren J., Pfeffer M., et al. Clinical features and contemporary management of patients with low and preserved ejection fraction heart failure: baseline characteristics of patients in the Candesartan in Heart Failure-Assessment of Reduction in Mortality and Morbidity (CHARM) programme. Eur J Heart Fail (2003) 5:261–270.[Abstract/Free Full Text]
  5. Curtis A.B., Gersh B.J., Corley S.D., et al. Clinical factors that influence response to treatment strategies in atrial fibrillation: the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study. Am Heart J (2005) 149:645–649.[CrossRef][Web of Science][Medline]
  6. Swedberg K., Olsson L.G., Charlesworth A., et al. Prognostic relevance of atrial fibrillation in patients with chronic heart failure on long-term treatment with beta-blockers: results from COMET. Eur Heart J (2005) 26:1303–1308.[Abstract/Free Full Text]
  7. DiMarco J.P., Flaker G., Waldo A.L., et al. Factors affecting bleeding risk during anticoagulant therapy in patients with atrial fibrillation: observations from the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study. Am Heart J (2005) 149:650–656.[CrossRef][Web of Science][Medline]
  8. Witte K.K., Desilva R., Chattopadhyay S., Ghosh J., Cleland J.G., Clark A.L. Are hematinic deficiencies the cause of anemia in chronic heart failure? Am Heart J (2004) 147:924–930.[CrossRef][Web of Science][Medline]
  9. Olsson S.B. Stroke prevention with the oral direct thrombin inhibitor ximelagatran compared with warfarin in patients with non-valvular atrial fibrillation (SPORTIF III): randomised controlled trial. Lancet (2003) 362:1691–1698.[CrossRef][Web of Science][Medline]
  10. Halperin J.L. Ximelagatran compared with warfarin for prevention of thromboembolism in patients with nonvalvular atrial fibrillation: rationale, objectives, and design of a pair of clinical studies and baseline patient characteristics (SPORTIF III and V). Am Heart J (2003) 146:431–438.[CrossRef][Web of Science][Medline]
  11. Albers G.W., Diener H.C., Frison L., et al. Ximelagatran vs warfarin for stroke prevention in patients with nonvalvular atrial fibrillation: a randomized trial. JAMA (2005) 293:690–698.[Abstract/Free Full Text]
  12. Diener H.C. Stroke prevention using the oral direct thrombin inhibitor ximelagatran in patients with non-valvular atrial fibrillation. Pooled analysis from the SPORTIF III and V studies. Cerebrovasc Dis (2006) 21:279–293.[CrossRef][Web of Science][Medline]
  13. Khand A.U., Cleland J.G., Deedwania P.C. Prevention of and medical therapy for atrial arrhythmias in heart failure. Heart Fail Rev (2002) 7:267–283.[CrossRef][Medline]
  14. van Walraven C., Hart R.G., Singer D.E., et al. Oral anticoagulants vs aspirin in nonvalvular atrial fibrillation: an individual patient meta-analysis. JAMA (2002) 288:2441–2448.[Abstract/Free Full Text]
  15. Lip G.Y., Gibbs C.R. Does heart failure confer a hypercoagulable state? Virchow's triad revisited. J Am Coll Cardiol (1999) 33:1424–1426.[Free Full Text]
  16. Cleland J.G. Anticoagulant and antiplatelet therapy in heart failure. Curr Opin Cardiol (1997) 12:276–287.[Web of Science][Medline]
  17. Cleland J.G., Massie B.M., Packer M. Sudden death in heart failure: vascular or electrical? Eur J Heart Fail (1999) 1:41–45.[Free Full Text]
  18. Horwich T.B., Patel J., MacLellan W.R., Fonarow G.C. Cardiac troponin I is associated with impaired hemodynamics, progressive left ventricular dysfunction, and increased mortality rates in advanced heart failure. Circulation (2003) 108:833–838.[Abstract/Free Full Text]
  19. Orn S., Cleland J.G., Romo M., Kjekshus J., Dickstein K. Recurrent infarction causes the most deaths following myocardial infarction with left ventricular dysfunction. Am J Med (2005) 118:752–758.[CrossRef][Web of Science][Medline]
  20. Uretsky B.F., Thygesen K., Armstrong P.W., et al. Acute coronary findings at autopsy in heart failure patients with sudden death: results from the assessment of treatment with lisinopril and survival (ATLAS) trial. Circulation (2000) 102:611–616.[Abstract/Free Full Text]
  21. Petersen P., Boysen G., Godtfredsen J., Andersen E.D., Andersen B. Placebo-controlled, randomised trial of warfarin and aspirin for prevention of thromboembolic complications in chronic atrial fibrillation. The Copenhagen AFASAK study. Lancet (1989) 1:175–179.[Web of Science][Medline]
  22. Shelton R.J., Clark A.L., Goode K., Rigby A.S., Cleland J.G. The diagnostic utility of N-Terminal pro-B-type Natriuretic for the detection of major strucural heart desease in patients with atrial fibrillation. Eur Heart J (2006) 27(19):2353–2361.[Abstract/Free Full Text]
  23. Cooper H.A., Dries D.L., Davis C.E., Shen Y.L., Domanski M.J. Diuretics and risk of arrhythmic death in patients with left ventricular dysfunction. Circulation (1999) 100:1311–1315.[Abstract/Free Full Text]
  24. Cleland J.G., Ghosh J., Freemantle N., et al. Clinical trials update and cumulative meta-analyses from the American College of Cardiology: WATCH, SCD-HeFT, DINAMIT, CASINO, INSPIRE, STRATUS-US, RIO-Lipids and cardiac resynchronisation therapy in heart failure. Eur J Heart Fail (2004) 6:501–508.[Abstract/Free Full Text]
  25. Cleland J.G., Findlay I., Jafri S., et al. The Warfarin/Aspirin Study in Heart failure (WASH): a randomized trial comparing antithrombotic strategies for patients with heart failure. Am Heart J (2004) 148:157–164.[CrossRef][Web of Science][Medline]
  26. Cleland J.G. Is aspirin "the weakest link" in cardiovascular prophylaxis? The surprising lack of evidence supporting the use of aspirin for cardiovascular disease. Prog Cardiovasc Dis (2002) 44:275–292.[CrossRef][Web of Science][Medline]
  27. Cleland J.G., Bulpitt C.J., Falk R.H., et al. Is aspirin safe for patients with heart failure? Br Heart J (1995) 74:215–219.[Free Full Text]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Eur Heart JHome page
A. Goette, A. Bukowska, D. Dobrev, J. Pfeiffenberger, H. Morawietz, D. Strugala, I. Wiswedel, F.-W. Rohl, C. Wolke, S. Bergmann, et al.
Acute atrial tachyarrhythmia induces angiotensin II type 1 receptor-mediated oxidative stress and microvascular flow abnormalities in the ventricles
Eur. Heart J., June 1, 2009; 30(11): 1411 - 1420.
[Abstract] [Full Text] [PDF]

Home page
 J Am Coll CardiolHome page
R. Nieuwlaat, L. W. Eurlings, J. G. Cleland, S. M. Cobbe, P. E. Vardas, A. Capucci, J. L. Lopez-Sendon, J. G. Meeder, Y. M. Pinto, and H. J.G.M. Crijns
Atrial fibrillation and heart failure in cardiology practice: reciprocal impact and combined management from the perspective of atrial fibrillation: results of the Euro Heart Survey on atrial fibrillation.
J. Am. Coll. Cardiol., May 5, 2009; 53(18): 1690 - 1698.
[Abstract] [Full Text] [PDF]

Home page
 Eur Heart JHome page
M. Rivero-Ayerza, W. Scholte op Reimer, M. Lenzen, D. A.M.J. Theuns, L. Jordaens, M. Komajda, F. Follath, K. Swedberg, and J. G.F. Cleland
New-onset atrial fibrillation is an independent predictor of in-hospital mortality in hospitalized heart failure patients: results of the EuroHeart Failure Survey
Eur. Heart J., July 1, 2008; 29(13): 1618 - 1624.
[Abstract] [Full Text] [PDF]

Home page
 Eur Heart JHome page
A. Torabi, J. G.F. Cleland, N. K. Khan, P. H. Loh, A. L. Clark, F. Alamgir, J. L. Caplin, A. S. Rigby, and K. Goode
The timing of development and subsequent clinical course of heart failure after a myocardial infarction
Eur. Heart J., April 1, 2008; 29(7): 859 - 870.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (4)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Cleland, J. G.F.
Right arrow Articles by Grind, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cleland, J. G.F.
Right arrow Articles by Grind, M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?