Response to letter to editor by Jolobe ‘Stroke risk might be specifically correlated with the burden of paroxysmal atrial fibrillation’

Dear Sir, I thank Dr Jolobe for his interest in our article [1]. I agree that it is unclear on what is the required ‘burden’ of atrial fibrillation (AF) necessary for stroke and thromboembolism, given that the number of daily AF episodes – as well as AF burden – can have great variability within (and between) patients [2]. This is also not unsurprising because of the general under-representation of paroxysmal AF patients in clinical trials. Also, it does depends on how hard we look for paroxysmal AF in patients who present with an acute stroke or thromboembolism, especially as paroxysms of AF are frequently asymptomatic [2, 3]. Indeed, the only study to investigate the relationship of AF burden to thromboembolism comes from the Italian AT500 Registry Investigators[4], where the adjusted risk of thromboembolism was 3.1-fold increased where device-detected AF episodes were >24 h during follow-up. However, this was a cohort of elderly patients with multiple comorbidities suffering from bradycardia and implanted with a pacemaker with antitachycardia pacing therapies, and further studies in ‘general’ populations of paroxysmal AF are still needed. Some insights into the pathophysiology of AF burden and thromboembolism have been obtained from surrogate indices of thrombogenesis, and data from the nonwarfarin arms of three trial cohorts have been published, as reviewed by Lip [2]. ‘Low’ AF burden does not equate to a ‘low’ stroke risk, in view of the variability in the number of paroxysms and arrhythmia burden in patients as well as the natural history of the development of AF which is influenced by age and comorbidities such as hypertension. Indeed, we have a one in four lifetime risk of developing AF [5], and a low AF burden in a particular patient with paroxysmal AF today does not mean that this would be the case over the next few months or years. Thus, paroxysmal AF patients should be given appropriate thromboprophylaxis and stroke risk stratification schema would apply equally to paroxysmal or permanent AF [6, 7]. The undisputable fact is that those patients with paroxysmal AF identified as being at ‘high-risk’ based on clinical or echocardiographic risk factors should not be denied oral anticoagulation. None declared.