Reducing Potential COVID-19 Exposure With Coronary Computed Tomography for Left Atrial Appendage Assessment: A Tool For Our Time

To the Editor,

The severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2) pandemic has been affecting Australia since January 2020 and currently the state of Victoria has emerged from lockdown for a third time. Aerosol generating procedures such as transoesophageal echocardiography (TOE) have been associated with a high risk of exposure to the operator and anaesthetic staff. Imaging Council guidelines released during the initial phase recommend delaying elective TOE for a minimum of 3 months [1]. However this is not a long-term solution if the pandemic becomes endemic.

Computed tomography assessment of the left atrial appendage (LAA) with delayed imaging has been shown to have a negative predictive value of 99% [2]. Delayed imaging involves performing a second scan after a short time delay (ranging from 30–180 secs), as it is difficult on the early phase scan to distinguish thrombus from sludge or spontaneous echo contrast. The impressive negative predictive value has been the basis of the Society of Cardiac Computed Tomography (SCCT) recommendation for replacing TOE with CT assessment where possible during the SARS-CoV-2 pandemic [3].

Our protocol was performed using Dual Source CT Scanner (Siemens SOMATOM Force, Siemens, Munich, Germany) with an anteroposterior (AP) topogram performed for anatomical localisation. A contrast timing bolus scan (10 mL of visipaque 320 with a 40 mL saline flush) assessed peak enhancement time within the ascending aorta. An initial electrocardiogram (ECG) gated helical scan was performed with a contrast injection of 60 mL of contrast at 5.5 mL/s followed by 40 mL of a 50% contrast 50% saline mix, and a further 40 mL saline flush at 5.0 mL/s. A second ECG gated helical scan followed with a timed delay of 60 seconds post initial scan, used to differentiate between pseudo thrombus and true thrombus. As both scans only need to include the LAA, the scan length is minimised and dose was reduced. Images were acquired in systole due to atrial fibrillation with rapid ventricular rate, however diastole can be used depending on heart rate. Scan start timing was calculated from the timing bolus, with images being acquired 3 seconds (5 secs would be used for high pitch FLASH scans) after peak enhancement in the ascending aorta. Reconstructed slice thickness is recommended at 0.6 mm an interval of 0.3 mm. A reconstruction kernel of Bv40 with an iterative reconstruction (Advanced Model Iterative Reconstruction [ADMIRE]) strength of 3 was used.

We performed three CT LAA scans with radiation doses of 1.6 mSv, 3 mSv and 20 mSv. The scans were successful in excluding LAA thrombus in two of the patients and the patients proceeded to direct current cardioversion without complication. The patient in whom an intracardiac thrombus could not be excluded proceeded to TOE and subsequent cardioversion.

Existing expertise reflecting the wide practice of computed tomography coronary angiography in Australia should mean that adoption of CT appendage assessment should not be cumbersome or burdensome. With the emergence of new variants of SARS-CoV-2 of increased infectivity, reduction in aerosol generating procedures can only be beneficial. We propose that CT assessment of the appendage becomes standard of care in the ongoing efforts to minimise exposure to COVID-19 in Australia and should persist in the post COVID-19 world.

Source of Funding

No external financial support was received.

Disclosures

The authors report no relevant disclosures.