Prediction and prevention of atrial fibrillation in patients with high blood pressure or history of hypertension

Abstract

In this issue of Journal of Hypertension, Verdecchia et al. [1] report on predictors for incident atrial fibrillation in 30 424 ONTARGET/TRANSCEND participants with vascular disease or complicated diabetes who were in sinus rhythm and averaged about 66 years of age at entry. A copy of the ECG was sent to the central office every time new atrial fibrillation was detected by the investigators. During a median follow-up period of 4.7 years, new atrial fibrillation occurred in 2092 patients, giving large statistical power to assess the predictors of this most common cardiac arrhythmia. Risk of atrial fibrillation increased with age, SBP, pulse pressure, left ventricular hypertrophy, BMI, serum creatinine, history of hypertension, history of coronary artery disease and history of cerebrovascular disease. After adjustment for BMI and other variables, the risk of atrial fibrillation also increased with hip circumference. New atrial fibrillation not only strongly portended an increased risk of congestive heart failure but also highly significantly predicted cardiovascular death. Risk of stroke was unaffected, whereas that of myocardial infarction was reduced. Patients with new atrial fibrillation were also more likely to receive vitamin K antagonists, statins and β-blockers than those in sinus rhythm. Undisputable strengths of this study [1] were the large sample size, the attention to careful follow-up of patients in the settings of a randomized clinical trial, the wide use of ECG in atrial fibrillation detection and the huge number of incident cases of atrial fibrillation. This population was rather well treated for their underlying hypertension at baseline, and most patients achieved rather low blood pressures during follow-up, perhaps in some fragile patients probably too low blood pressure [2]. It is remarkable that, despite this intensive therapeutic approach, high blood pressure and markers of hypertensive disease had such an impact on the incidence of atrial fibrillation. Thus, left ventricular hypertrophy, serum creatinine, history of hypertension and history of cerebrovascular disease were highly significant independent predictors, as were BMI and history of coronary heart disease. The present ONTARGET/TRANSCEND analysis [1] is a massive documentation of the connection between hypertension, its end-organ damage and associated risk factors and the risk of developing atrial fibrillation. This relationship between history of hypertension and risk of atrial fibrillation persisted despite confounding from the extensive vascular disease or complicated diabetes present in many ONTARGET/TRANSCEND participants or, perhaps, these diseases even escalated the relationship. Due to its high prevalence in the population, hypertension is by far the most prevalent risk factor for the development of atrial fibrillation [3–5]. Recent studies in women [6] and in men [7] show that blood pressures even in the upper normal range are associated with an increased risk of atrial fibrillation. The dominating role of high blood pressure in the causes, pathophysiology, prediction, diagnosis and treatment of atrial fibrillation was recently extensively reviewed by the European Society of Hypertension [8]. Most remarkable is the observation that up to about 90% of patients with atrial fibrillation participating in some of the recent large randomized clinical trials of new medications in atrial fibrillation had a history of hypertension [8], suggesting atrial fibrillation is in most cases a typical complication of hypertension, even more so than stroke and heart failure. It is not entirely surprising that left ventricular hypertrophy is a strong predictor of atrial fibrillation. In the LIFE study, a large trial in patients with left ventricular hypertrophy (n = 9193), blood pressures, Cornell voltage–duration product, left ventricular hypertrophy on ECG, age and the male sex were the predictors of atrial fibrillation [9]. In LIFE, also lower heart rate on the ECG during treatment was highly beneficial [10] and there was clearly less incident atrial fibrillation with regression of left ventricular hypertrophy [11]. A recent and very detailed analysis [12] shows that in patients with left ventricular hypertrophy, pulse pressure was equivalent to SBP and DBP combined in predicting new onset atrial fibrillation, but when forced into the same statistical model, the pulse pressure was by far the strongest predictor of new atrial fibrillation among the various blood pressure components. This strong association with pulse pressure suggests that in advanced hypertensive disease, a key mechanism for promoting the electrical instability leading to atrial fibrillation may consist in the elevated arterial stiffness coupled with increased afterload causing atrial walls stretch with atrial chamber dilation and increased pressure into the pulmonary veins. The presence of left ventricular hypertrophy by ECG, age and coronary heart disease were similarly significant confounding covariates in another large sample (n = 15 300) of hypertensive people participating in the VALUE trial, which also included proper serial ECG detection of incident atrial fibrillation [13]. Clearly, as now also seen in this analysis of ONTARGET/TRANSCEND [1], incident atrial fibrillation quite strongly portended an increased risk of congestive heart failure [14], which may be explained by the fact that people with more advanced cardiac disease cannot sustain the impact of losing 20–25% of their cardiac output. There is further wisdom from ONTARGET. This huge clinical trial was organized in order to thoroughly investigate the impact of inhibiting the renin–angiotensin system with ramipril, telmisartan or the combination. The main publication reported that 570 (6.9%), 550 (6.7%) and 537 (6.5%) patients developed new atrial fibrillation in each of the three treatment arms of the trial [15]. In the light of the high risk of the ONTARGET population, the incidence of new atrial fibrillation may seem to be quite low. However, ONTARGET patients were already intensely treated for their high blood pressures and addition of the combination therapy on the background of this intense therapy may have been harmful for fragile patients [2], and this may have counterbalanced the possible benefit of dual blockade of the renin–angiotensin system with respect to prevention of incident atrial fibrillation. Blocking the renin–angiotensin system does not appear to readily work in preventing recurrent atrial fibrillation [16,17], unless combined with amiodarone [18]. However, in the treatment of less severe or complicated hypertension, treatment with losartan and valsartan has been shown to be particularly effective compared with treatment with atenolol [19] and amlodipine [13] in preventing atrial fibrillation, possibly due to greater regression of left ventricular hypertrophy with these therapies [11]. Thus, prespecified secondary analyses of two large prospective randomized clinical trials with double-blinded design and use of incident atrial fibrillation taken from serial ECGs as an endpoint clearly showed benefits of angiotensin receptor blockade over the most commonly used β-blocker and calcium channel blocker prescribed for hypertension and cardiovascular disease in the world [13,18]. This qualifies the evidence in favor choosing angiotensin receptor blockers as class 1 level A, as in the guidelines of the European Society of Cardiology [20], and is a strong plea for the extensive use of these compounds in the treatment of hypertension and the prevention of atrial fibrillation. ACKNOWLEDGEMENTS Conflicts of interest The authors were in the leadership of the LIFE and VALUE studies, supported by Merck and Novartis.