Bidirectional Embolism in Atrial Septal Defect

To the Editor: Paradoxic embolism is considered to be an occasional cause of stroke in patients with atrial septal defect (ASD),1 although the recent finding of a greater prevalence of atrial shunt in patients with migraine with aura and stroke than in those without stroke suggests that paradoxic embolism through an ASD might be a more common cause of stroke.2 Overall, paradoxic embolism is considered to be responsible for approximately 2% of all cases of systemic arterial emboli.3 We describe a case of bidirectional embolism in an elderly patient with recent transient aphasia, ASD, and pulmonary hypertension. An 80-year-old woman with recently diagnosed “congestive heart failure in patients with ASD ostium secundum complicated by pulmonary hypertension” underwent a transthoracic echocardiography in the context of an examination for worsening dyspnea. Her history also included one episode of transient aphasia lasting approximately 20 hours in the week before admission, with cerebral magnetic resonance imaging revealing mainly periventricular white matter lesions, which were likely ischemic or hypoxemic in origin. On physical examination, the patient appeared to be acutely sick, tachypneic at rest, and cachectic. She had a prominent jugular pulse, a loud pulmonary component of second sound, a systolic murmur on the right border of the sternum, and rales at both pulmonary bases; her systemic blood pressure was 100/70 mmHg. The electrocardiogram showed atrial fibrillation and right bundle block. Arterial blood gases were as follows: arterial pressure of oxygen, 52.1 mmHg; arterial pressure of carbon dioxide, 48.8 mmHg; pH, 7.43, and oxygen saturation, 87.3%. The echocardiography showed a dilated and hypokinetic right ventricle, severe pulmonary hypertension (100 mmHg), and a large ASD with bidirectional shunt. Several spots floating in the atria and the right ventricle were detected (Figure 1A–D; see video supplement). The spots were moving through the ASD in both directions according to the respiratory cycle. Transesophageal echocardiography confirmed the bidirectional embolism through the ASD and showed parietal thrombi in the right atrium and the left atrial appendage (Figure 1E, F). The absence of a smoke-like effect excluded the presence of spontaneous echo contrast. The D-dimer value was 2,093 ng/mL (normal range 0–500 ng/mL). An echocolordoppler excluded the presence of carotid or vertebral stenosis. Therapy with enoxaparin, sildenafil, and inhaled iloprost was started. Bilevel positive airway pressure was also used because of respiratory muscle exhaustion. A control echocardiography 1 week later found the pulmonary pressure to be 80 mmHg. The patient's general condition improved (from functional New York Heart Association Class IV to Class III), and her arterial blood pressure rose to 125/80 mmHg. Transthoracic echocardiographic images showing embolic spots crossing the atrial septal defect. In the upper panels, spot 2 is crossing the septum from the left (A) to the right atrium (B). The middle panels show spots 3 and 4, which are going from the right (C) to the left (D) atrium (see video supplement). In the lower panels, transesophageal echocardiography showing parietal thrombi in right atrium (E) (arrows) and left atrial appendage (F) (arrows). This case gives support to paradoxical embolism as a cause of transient ischemic attack, although the parietal left atrial thrombi might be an alternative source of emboli. Alternatively, severe systemic hypotension and hypoxemia seem an unlikely explanation of a localized neurological deficit such as transient aphasia. Congenital atrial abnormalities have been studied in relation to the risk of stroke. Of these, ASD, atrial septal aneurysm, and eustachian valve anatomy received particular attention,4 although even ASD causing an important shunt can remain unrecognized until old age because hemodynamic decompensation secondary mainly to pulmonary hypertension may require several decades to develop. Unrecognized ASD with late development of symptoms has been previously described in elderly patients.5 Accordingly, the age at which nonneurological symptoms, mainly exertional dyspnea and fatigue, first appear is highly variable and not exclusively related to the size of the shunt.1 As a consequence, ASD may be detected on the occasion of a diagnostic examination for unexplained neurological deficits, although new-onset heart failure, abnormal vascular chest x-ray, or unexplained hypoxemia (due to venous blood admixture) are more likely clues to the diagnosis. In our patient, ASD was the site of bidirectional embolization, likely due to the severe pulmonary hypertension; embolization was leftward (i.e., paradoxic) during exhalation, when right atrial pressure physiologically increases because of the rise in intrathoracic pressure. This in vivo demonstration of paradoxic embolism makes the case interesting. Furthermore, the relationship between ASD and cerebral ischemia, which can be considered to be established in young-adult subjects, is not obvious in elderly people.4,6–10 In conclusion, the finding of paradoxic embolism through an ASD lends support to the view that such a mechanism could underlie otherwise unexplained cerebral ischemia at any age. It also suggests that patients with pulmonary hypertension should be screened for atrial or ventricular right to left shunt driven by increased pressures in the right heart cavities. Conflict of Interest: The editor in chief has reviewed the personal and financial checklist provided by the authors and determined that none of the authors have any conflicts related to this letter. Author Contributions: Girolamo Catapano-Minotti, Mario Baldi, Rita Naio, and Raffaele Antonelli Incalzi participated in case study, acquisition and interpretation of data, and preparation of manuscript. Andrea Corsonello participated in interpretation of data and preparation of manuscript. Sponsor's Role: None.