Author + information
- Received January 26, 2004
- Accepted June 9, 2004
- Published online March 1, 2005.
- Kristian Wachtell, MD, PhD*,†,* (, )
- Björn Hornestam, MD, PhD‡,
- Mika Lehto, MD§,
- David J. Slotwiner, MD, FACC†,
- Eva Gerdts, MD, PhD∥,
- Michael H. Olsen, MD, PhD*,
- Peter Aurup, MD¶,
- Björn Dahlöf, MD, PhD, FACC‡,
- Hans Ibsen, MD*,
- Stevo Julius, MD, FACC#,
- Sverre E. Kjeldsen, MD, PhD, FACC**,
- Lars H. Lindholm, MD††,
- Markku S. Nieminen, MD§,
- Jens Rokkedal, MD* and
- Richard B. Devereux, MD, FACC†
- ↵*Reprint requests and correspondence:
Dr. Kristian Wachtell, Rigshospitalet, Department of Medicine B2142, 9 Blegdamsvej, DK-2100 Copenhagen, Denmark
Objectives We assessed the impact of antihypertensive treatment in hypertensive patients with electrocardiographic (ECG) left ventricular (LV) hypertrophy and a history of atrial fibrillation (AF).
Background Optimal treatment of hypertensive patients with AF to reduce the risk of cardiovascular morbidity and mortality remains unclear.
Methods As part of the Losartan Intervention For End point reduction in hypertension (LIFE) study, 342 hypertensive patients with AF and LV hypertrophy were assigned to losartan- or atenolol-based therapy for 1,471 patient-years of follow-up.
Results The primary composite end point (cardiovascular mortality, stroke, and myocardial infarction) occurred in 36 patients in the losartan group versus 67 in the atenolol group (hazard ratio [HR] = 0.58, 95% confidence interval [CI] 0.39 to 0.88, p = 0.009). Cardiovascular deaths occurred in 20 versus 38 patients in the losartan and atenolol groups, respectively (HR = 0.58, 95% CI 0.33 to 0.99, p = 0.048). Stroke occurred in 18 versus 38 patients (HR = 0.55, 95% CI 0.31 to 0.97, p = 0.039), and myocardial infarction in 11 versus 8 patients (p = NS). Losartan-based treatment led to trends toward lower all-cause mortality (30 vs. 49, HR = 0.67, 95% CI 0.42 to 1.06, p = 0.090) and fewer pacemaker implantations (5 vs. 15, p = 0.065), whereas hospitalization for heart failure took place in 15 versus 26 patients and sudden cardiac death in 9 versus 17, respectively (both p = NS). The benefit of losartan was greater in patients with AF than those with sinus rhythm for the primary composite end point (p = 0.019) and cardiovascular mortality (p = 0.039).
Conclusions Losartan is more effective than atenolol-based therapy in reducing the risk of the primary composite end point of cardiovascular morbidity and mortality as well as stroke and cardiovascular death in hypertensive patients with ECG LV hypertrophy and AF.
Atrial fibrillation and flutter (AF) increase the risk of cardiovascular morbidity and mortality (1,2), and hypertensive patients have up to a 42% increased risk of developing AF (1,3). Because many hypertensive patients develop AF, the addition of AF to hypertension contributes to increased rates of cardiovascular morbidity and mortality in hypertension. Traditionally, restoration of sinus rhythm by direct current conversion has been attempted in patients with AF (4), although this treatment remains controversial. A recent study showed that management of AF with a rhythm-control strategy offered no survival advantage over the rate-control strategy (5). Therefore, combined use of digoxin and either beta-blockers or calcium antagonists to treat hypertension and achieve rate control may be increasingly preferred by clinicians (4). However, no data are available to determine whether traditional beta-blocker–based therapy or direct inhibition of the renin-angiotensin system reduces cardiovascular mortality and morbidity more in hypertensive patients with AF. As previously described (6–8), the Losartan Intervention For End point reduction in hypertension (LIFE) study primary hypothesis was that selective angiotensin-II type 1 receptor blockade with losartan would be more effective than beta-blockade with atenolol in reducing cardiovascular morbidity and mortality in hypertensive patients with electrocardiographic (ECG) left ventricular (LV) hypertrophy. Therefore, we analyzed the outcome in the subgroup of hypertensive patients with ECG LV hypertrophy (n = 342) who had a history of AF or electrocardiographically verified AF at the start of the LIFE study to evaluate the treatment effect of a specific angiotensin-II receptor versus beta-blockade on cardiovascular morbidity and mortality in hypertensive patients with ECG LV hypertrophy.
The LIFE study was a prospective, randomized, double-blind, parallel-group study with a double-dummy technique. The primary objective and main outcome, as well as the complete study protocol with study design, organization, clinical measures, end point definitions, exclusion criteria, basis for choice of comparative agents, statistical considerations, and baseline characteristics, have been previously published (6–8).
Patients age 55 to 80 years, having previously treated or untreated hypertension and ECG LV hypertrophy by either Cornell voltage-duration or Sokolow-Lyon voltage criteria (9), were randomized to initial therapy with 50 mg/day losartan or atenolol after one to two weeks of placebo if they had a sitting systolic blood pressure of 160 to 200 mm Hg and/or diastolic blood pressure of 95 to 115 mm Hg. In both groups, hydrochlorothiazide was added in the case of insufficient pressure lowering. Thereafter the study drug was increased to 100 mg/day and supplemented with additional antihypertensive therapy in order to reach a target blood pressure of <140/90 mm Hg. Patients were enrolled from June 1995 to May 1997 and were followed for four years or longer. Of 342 LIFE study participants (3.7% of 9,193) who had either electrocardiographically documented AF or a history of AF (n = 324) or atrial flutter (n = 18) reported by the investigator at baseline, 157 were randomized to losartan and 185 to atenolol. Centralized ECG reading confirmed that 135 patients (40%) had persistent or permanent AF documented by electrocardiography. The care of patients' AF was left to the discretion of the physician.
All end points were analyzed using the intention-to-treat approach, all randomized patients were included in their randomized treatment group, and all available follow-up data were included from randomization through the study termination date. Only end points confirmed by the Endpoint Committee were included in analyses. Patients with multiple end points were counted as having had an event in all relevant end point analyses; however, only the first event in a specific category counted in any individual analysis. The difference between treatment groups with respect to clinical events was assessed by Cox regression models with the degree of LV hypertrophy (as measured by both Cornell voltage-duration product and Sokolow-Lyon voltage) and the Framingham risk score (10) at baseline and difference in blood pressure during treatment as covariates. This adjusted analysis was chosen a priori primarily to account for any difference in key risk predictors at baseline; additional adjustments based on findings in the present study are described. The risk reduction for losartan versus atenolol was calculated as: 100·(1 − relative risk). Event rates over time are presented as Kaplan-Meier curves: the numbers below the curves represent the number of event-free patients remaining in follow-up at the corresponding time point. Differences in baseline characteristics were assessed by Student ttest and chi-square for categorical variables. Tests were performed at two-sided 5% significance levels. For further details, see Dahlöf et al. (8).
History of AF versus sinus rhythm
Demographic characteristics of patients with or without a history of AF at enrollment in the LIFE study are compared in Table 1.Heart rate was similar at baseline and during treatment in patients with a history of AF compared with patients in sinus rhythm (p = NS).
Patients with a history of AF had, compared with those without, higher rates of cardiovascular and all-cause mortality, fatal and non-fatal stroke, heart failure (HF), revascularization, and sudden cardiac death, but statistically similar rates of myocardial infarction (MI) and hospitalization for angina pectoris (Table 2).
Treatment of patients with a history of AF
Patients with a history of AF assigned to losartan- or atenolol-based treatment had comparable demographic characteristics (Table 1). During the study medications were titrated similarly in the two treatment arms. At end point or termination of follow-up 8.3% and 11.4% of losartan- and atenolol-treated patients, respectively, received 50 mg of study drug alone; 12.7% and 14.5%, respectively, received 50 mg of study drug plus hydrochlorothiazide or other medications; and 43% and 31%, respectively, received 100 mg of the study drug with or without hydrochlorothiazide or additional medications. Furthermore, hydrochlorothiazide was administered in 61% of study duration in losartan-treated and 52% of study duration in atenolol-treated patients (p = NS).
Patients with a history of AF had a similar high prevalence of use of another antihypertensive therapy (e.g., centrally acting and/or non-dihydropyridine calcium channel blocker) and received similar concomitant treatment with other rate-controlling, antiarrhythmic and lipid-lowering drugs during the study (Table 3).Most patients adhered to the treatment regimen, and crossover treatment occurred mainly in patients who discontinued study medication. Among patients originally assigned to losartan, 27.6% crossed over to beta-blockade, and among those originally assigned to atenolol, 34.9% crossed over to angiotensin-converting enzyme inhibition/angiotensin-II receptor blockade by the study's end. Patients who discontinued losartan were no more likely to receive open-label beta-blockade than patients who discontinued atenolol (27.6% vs. 31.7%, p = NS). Open-label combination therapy of beta-blocker and angiotensin-converting enzyme inhibition/angiotensin-II receptor blockade during the study was seen in 14.6% and 21.6%, respectively, among losartan- and atenolol-assigned patients (p = 0.123).
Systolic, diastolic, and mean blood pressures were reduced substantially in both losartan- and atenolol-treated patients (Fig. 1).Average blood pressure at the last visit before a primary end point, or at the end of follow-up, was 145/81 and 147/79 mm Hg in the losartan and atenolol groups, respectively, showing reductions of 31/17 and 26/16 mm Hg. The average reduction in heart rate during the first 12 months of treatment was significantly greater in atenolol-treated patients (9.9 beats/min vs. 3.5 beats/min, p < 0.001), but very similar at study end (Fig. 2).There was a trend toward higher serum potassium during treatment in patients with a history of AF compared with those in sinus rhythm (4.21 ± 0.41 mmol/l vs. 4.17 ± 0.39 mmol/l, p = 0.06). Among patients with a history of AF there was no difference in serum potassium during treatment between losartan-treated and atenolol-treated groups (4.20 ± 0.43 mmol/l vs. 4.22 ± 0.40 mmol/l, p = NS).
Patients with a history of AF on losartan treatment had a significantly lower rate of the primary composite end point than those treated with atenolol during 1,471 patient-years of follow-up (Table 4,Fig. 3).Losartan-treated patients also had lower rates of cardiovascular death and stroke, with no difference in MI rate. The treatment effects on hemorrhagic or ischemic stroke were similar (data not shown). There was no evidence of interaction between treatment and gender for any pre-specified end point (all p > 0.144). A pacemaker was implanted in 15 (8.1%) atenolol-treated and 5 (3.2%) losartan-treated patients, respectively, during the study (p = 0.065). Furthermore, losartan tended to reduce the prevalence of electrocardiographically documented AF by 14% at the annual visits during the following 4.8 years observation (losartan 50.3% vs. atenolol 55.7%, hazard ratio [HR] = 0.86, 95% confidence interval [CI] 0.64 to 1.15, p = 0.30).
Comparing patients with or without AF at baseline and using tests for interaction in Cox regression models with Framingham risk score and measures of ECG LV mass as covariates, the benefit of losartan was greater in patients with a history of AF than in other patients for both the primary composite end point (p = 0.019 for interaction) and for cardiovascular mortality (p = 0.039 for interaction).
Alternative adjustment for differences in the prevalence of peripheral vascular disease, diabetes, and isolated systolic hypertension instead of Framingham Risk Score yielded HRs associated with losartan treatment of 0.61 for the primary composite end point (95% CI 0.41 to 0.92, p = 0.019), 0.64 for cardiovascular mortality (95% CI 0.37 to 1.11, p = 0.114), 0.52 for stroke (95% CI 0.29 to 0. 92, p = 0.024), 1.70 for MI (95% CI 0.68 to 4.28, p = 0.260), and 0.74 for total mortality (95% CI 0.47 to 1.18, p = 0.206). Other secondary end points did not attain statistical significant difference (data not shown).
Among patients with a history of AF but without clinically recognized diabetes or coronary, cerebral, or peripheral vascular disease at enrollment in the LIFE study, the composite end point occurred in 10 of 77 patients in the losartan group versus 22 of 82 in the atenolol group (26.3 vs. 62.8 per 1,000 patient-years; HR = 0.57, 95% CI 0.38 to 0.87, p = 0.008). The secondary end point of stroke occurred in 5 versus 17 patients (13.1 vs. 47.1 per 1,000 patient-years) in the losartan and atenolol groups, respectively (HR = 0.52, 95% CI 0.29 to 0.92, p = 0.025). A parallel albeit non-significant trend was seen for cardiovascular death (12.8 vs. 25.7 deaths per 1,000 patient-years; HR = 0.60, 95% CI 0.35 to 1.03, p = 0.06), but not for MI or other secondary end points (data not shown).
The present study extends knowledge of AF and its treatment in several regards. First, we show that AF is associated with increased cardiovascular morbidity and mortality in hypertensive patients at high risk because of the presence of LV hypertrophy, as previously shown in predominately normotensive populations (11). Second, the present results provide the first evidence that greater prognostic benefit is provided to hypertensive patients with AF by one versus another medication with similar blood pressure lowering, building on previous evidence of the benefits of effective blood pressure control (12–14). The rate of the composite end point in the LIFE study, as well as the rates of cardiovascular mortality and stroke, was significantly lower with losartan- compared to atenolol-based therapy, beginning at one year of treatment. Third, the present study is the first to suggest that losartan is better than atenolol in reducing recurrence of AF in patients with a history of AF. Given the frequent occurrence of reduced exercise tolerance and increased dyspnea in AF compared with sinus rhythm, keeping patients in sinus rhythm is often a clinical priority.
Our study did not show reductions of MI and hospitalizations for angina pectoris or HF with losartan in hypertensive patients with a history of AF. However, beta-blockade is a well-documented treatment for ischemic heart disease and has shown substantial effect on morbidity and mortality in patients with angina pectoris, previous MI, and HF. A study of patients with hypertension and HF showed a 50% reduction in sudden death and a 30% reduction in hospitalizations for HF with metoprolol compared with placebo treatment (15) on top of standard treatments, including angiotensin-converting enzyme inhibition. Furthermore, beta-blocker use has been shown to independently reduce the risk of pump failure death in patients with AF (16). That treatment with losartan did as well as atenolol in MI and hospitalization for angina and HF is reassuring because of the high risk of coronary events in hypertensive patients with LV hypertrophy.
Another important observation of this study is the fact that beta-blocker treatment did not seem to reduce the risk of sudden cardiac death compared with losartan treatment. This is surprising, as many regard beta-blocker therapy as a first-line intervention for prevention of sudden cardiac death (17). In the merged European Myocardial Infarct Amiodarone Trial (EMIAT) and Canadian Amiodarone Myocardial Infarction Trial (CAMIAT) databases, concomitant use of beta-blockers and amiodarone provided a survival benefit, especially for arrhythmic cardiac death (18). In our study, even though the difference in sudden cardiac death did not attain statistical significance, there was a 48% risk reduction favoring the losartan-treated patients. Furthermore, beta-blocker–treated patients seemed to need increased pacemaker implantation even though patients at baseline had heart rates averaging 75 beats/min. Losartan also reduced heart rate, albeit less than atenolol, with less heart rate difference at the end of study (Fig. 2). This might be due to blockade of sympathetic nervous system activation by renin-angiotensin system blockade by losartan.
One explanation for the added effects by losartan on cardiovascular morbidity and mortality in this hypertensive population with LV hypertrophy could be the parallel effects of losartan on regression of atrial and ventricular hypertrophy. We have recently shown in the LIFE echocardiography substudy that patients with LV hypertrophy also exhibit increased left atrial size and hypertrophy (19), which have been associated with increased risk of stroke in population of normotensive and hypertensive adults (20). Further analyses from the LIFE echocardiography study will investigate whether left atrial diameter decreases with reduction in LV mass during treatment and whether this is related to a reduction in subsequent events.
Another possible explanation for reduced vascular events with losartan treatment could be greater effects on arterial atherothrombosis or on coagulation system activation in patients with AF. However, the fact that the benefit of losartan was not limited to high-risk patients with either vascular disease or diabetes, but also occurred in patients with hypertension and AF alone, suggests that the benefits of losartan in patients with AF are due to prevention of vascular disease as well as reduced progression of established arterial disease.
Furthermore, our study does not suggest high serum potassium as the cause of increased cardiovascular morbidity and mortality, because our hypertensive patients with AF treated with losartan had a slightly higher serum potassium level than those treated with atenolol.
The participants in the LIFE study were selected for hypertension and ECG LV hypertrophy but also for lack of current need for atenolol, losartan, or angiotensin-converting enzyme inhibitors, or known intolerance to primary study treatment. Furthermore, in view of the relatively low proportion of patients treated with non-dihydropyridine calcium blockers, most patients in the present study could be rate-controlled by digoxin alone or by no antiarrhythmic treatment during the study. However, in view of the large number of hypertensive patients age 55 to 80 years with LV hypertrophy meeting the LIFE study blood pressure criteria, estimated at 3.9 million in the U.S. and 4.8 million in the European Union (21,22), a considerable number of hypertensive patients with AF might have their risk of cardiovascular events and stroke reduced by losartan therapy.
Although the analysis of AF was not pre-specified in the LIFE study analysis plan from 1995, evaluation of treatment effects in the subgroup of patients with AF at baseline was a planned secondary analysis before termination (September 2001) and unblinding. Furthermore, since the sub-study population was recruited for hypertension and ECG LV hypertrophy and not for AF, balanced randomization is not guaranteed.
Because outcomes were analyzed by the intention-to-treat principle and without restriction after study drug discontinuation, open-label use of angiotensin-II receptor blocker/angiotensin-converting enzyme inhibitor/beta-blocker may have diminished the difference between the groups.
Losartan-based antihypertensive therapy was more effective than an atenolol-based regimen in reducing the risk of the primary composite end point of cardiovascular morbidity and cardiovascular mortality as well as the secondary end points of stroke and cardiovascular death in hypertensive patients with ECG LV hypertrophy and a history of AF. Hypertensive patients with AF who do not need beta-blockade for heart rate control seem to benefit more from losartan-based treatment than from conventional antihypertensive and anti-arrhythmic treatment.
We are indebted to Steven M. Snapinn for database management and Sigrid Helle Berg for her dedicated work with the LIFE study.
Drs. Wachtell, Gerdts, Olsen, Dahlöf, Ibsen, Kjeldsen, Nieminen, Rokkedal, and Devereux have received support for research from Merck and Co., Inc. Drs. Wachtell, Olsen, Dahlöf, Ibsen, Julius, Kjeldsen, Lindholm, Nieminen, and Devereux have received occasional speaking and consulting honoraria from Merck and Co., Inc., and Dr. Aurup was an employee of Merck and Co.
- Abbreviations and acronyms
- atrial fibrillation
- confidence interval
- heart failure
- hazard ratio
- Losartan Intervention For End point reduction in hypertension study
- left ventricular
- myocardial infarction
- Received January 26, 2004.
- Accepted June 9, 2004.
- American College of Cardiology Foundation
- Benjamin E.J.,
- Wolf P.A.,
- D'Agostino R.B.,
- et al.
- Fuster V.,
- Ryden L.E.,
- Asinger R.W.,
- et al.
- Dahlöf B.,
- Devereux R.B.,
- de Faire U.,
- et al.
- Okin P.M.,
- Devereux R.B.,
- Jern S.,
- et al.
- Anderson K.M.,
- Wilson P.W.,
- Odell P.M.,
- Kannel W.B.
- Tsang T.S.,
- Petty G.W.,
- Barnes M.E.,
- et al.
- Curb J.D.,
- Pressel S.L.,
- Cutler J.A.,
- et al.
- Kjeldsen S.E.,
- Westheim A.,
- Os I.
- Dries D.L.,
- Exner D.V.,
- Gersh B.J.,
- et al.
- Boutitie F.,
- Boissel J.P.,
- Connolly S.J.,
- et al.
- Benjamin E.J.,
- D'Agostino R.B.,
- Belanger A.J.,
- Wolf P.A.,
- Levy D.
- ↵Centers for Disease Control and Prevention. NHANES 1999–2000 public data release file documentation. Available at: http://www.cdc.gov/nchs/about/major/nhanes/currentnhanes.htm2004. Accessed January 15, 2004.