The role of atrial septal abnormalities in cryptogenic stroke — still questionable?

An atrial septal aneurysm consists of redundant atrial septal tissue bulging into the right or the left atrium, sometimes even oscillating between both atria. Different definitions for the cut-off point between a mobile and an aneurysmatic atrial septum have been used, which may partly contribute to the range of atrial septal aneurysm prevalence in various studies in recent years (2%–9%). However, there is general agreement in all definitions that the basal width of an atrial septal aneurysm should be more than 15 mm and the excursion of the aneurysm beyond the plane of the residual atrial septum should be at least 10 or 15 mm. Atrial septal aneurysm has been found in about 1% of consecutive autopsies and in up to 4·9% of patients undergoing transoesophageal echocardiography for reasons other than a search for sources of emboli. In contrast, a large number of studies have described a significantly higher percentage of atrial septal aneurysm between 4% and 15% in pre-selected patient populations with otherwise unexplained transient ischaemic attack and stroke. Most of these studies assessing the potential role of atrial septal aneurysm as a source of emboli have compared the prevalence of atrial septal aneurysm in patients undergoing transoesophageal echocardiography after an embolic event with the frequency in patients in whom transoesophageal echocardiography was performed for other clinical indications. However, these highly selected control groups may have led to a bias and an over-estimation of the true prevalence of atrial septal aneurysm in an unselected population. A recent study from Agmon et al. similarly demonstrated a higher percentage of atrial septal aneurysm in patients with otherwise unclear cerebral embolism (7·9%) compared to the prevalence in a large matched population-based and thus not pre-selected control group (2·2% of subjects within the SPARC study). In the present multicentre study Mattioli and co-workers analysed 245 patients with an otherwise unexplained cerebral event (transient ischaemic attack or stroke) and compared the prevalence of atrial septal anomalies to an ageand sex-matched control group of 316 subjects. The prevalence of atrial septal aneurysm both in the patient group (27·7%) and in the controls (9·9%) was nearly four times higher than in the latest study of Agmon See page 261 for the article to which this Editorial refers