To the Editor:

I appreciate the comments by Dr Cotter and colleagues about the improved sensitivity of transthoracic echocardiography (TTE) for the diagnosis of a patent foramen ovale (PFO). Having been among those who first applied contrast TTE to the study of the very relationship between PFO and ischemic stroke, I am not about to underestimate its importance. In my review, I underscored how the lower sensitivity of TTE compared with transesophageal echocardiography (TEE) was due mainly to patients with poor acoustic windows or very small shunts. The introduction of harmonic imaging has indeed improved the sensitivity of TTE for PFO detection, although at the possible expense of its specificity. As with contrast TEE, a rigorous methodology in performing the test, including image optimization and the performance of multiple contrast injections, is necessary and may be as important as the improved image quality. The reports mentioned by Cotter et al, in which TTE and also transcranial Doppler detected some PFOs that were not detected by TEE, can only be explained as reflecting an inadequate performance of the Valsalva maneuver during TEE, possibly due to patient oversedation. This is a circumstance that is acknowledged in my review and points at problems in the performance of the TEE rather than at an imperfect sensitivity of the test.

I do not agree with the concept of a “paradigm shift” in the detection of PFO mentioned by Cotter et al. A paradigm shift occurs whenever medical practice changes as the result of newly acquired information. I do not believe that the improvements in the sensitivity of TTE for PFO detection fall into this category. The performance of TEE is still needed for the direct visualization of a PFO, the evaluation of the presence of one or more sites of shunting, the accurate distinction of interatrial shunting from transpulmonary recirculation, and the evaluation of the anatomic suitability for transcatheter closure of a PFO (ie, the existence of a suitable rim for the deployment of a closure device); in other words, TEE is needed whenever an accurate diagnostic evaluation is needed to aid in the therapeutic decision-making process. In the conclusion of my review, I recommended that TTE be used as a screening test and that TEE be reserved for patients in whom additional assessment is necessary, such as to guide further therapeutic steps or evaluate their results. I believe that this recommendation represents the very acknowledgment of the complementary roles of the two diagnostic techniques also suggested by Cotter et al.