To the Editor:
We read with interest the comprehensive review article by Dr Di Tullio on the detection and relevance in stroke of patent foramen ovale. We would like to draw attention to the discussion of the appropriate investigation for the presence of a right-to-left shunt. The author notes transesophageal echocardiography (TEE) to be the gold standard investigation with a sensitivity and specificity of up to 100%. He states that contrast transthoracic echocardiography (TTE) has a lower sensitivity on the order of 50% to 60% when compared with TEE; this statement is accompanied by 8 references to studies comparing TEE with TTE or transcranial Doppler.
All of the references provided by the author are from the years 1991 to 1994, and indeed these and other articles from that time reached similar conclusions confirming the superiority of contrast TEE with sensitivities ranging from 37% to 63% for TTE compared with TEE. It was therefore appropriate at that time that TEE be considered the gold standard.
However, since then there has been a paradigm shift with technical advances in TTE. This, in combination with the ease in performing Valsalva with TTE, now ensures at least as good sensitivity and specificity with bubble-contrast TTE when compared with TEE in the detection of right-to-left shunts. This has been verified more recently by 4 studies that compared contrast TEE with TTE using harmonic imaging. These studies report the sensitivity of TTE when compared with TEE to be much improved with harmonic imaging, ranging from 66% to 100%. In the most recent of these studies, Daniëls and colleagues evaluated 256 patients and found 6 of 46 large shunts detected with TTE not to be detected by TEE. It has also been noted for large shunts detected with transcranial Doppler to be missed with TEE. Our experience also suggests an extremely high sensitivity and specificity with bubble-contrast TTE, with detection in 75% of young patients with cryptogenic stroke (PE Cotter MB 2010 unpublished).
We fully agree that TEE is superior in the assessment of inter-atrial septum with respect to the diagnosis of an atrial septal aneurysm. Perhaps the most practical point is that the 2 tests are complementary, and we should avoid being didactic with respect to which investigation should be used to the exclusion of all other tests.