Almost half the patients implanted with a transcatheter aortic valve replacement (TAVR) device received the wrong valve size in a study that compared oversizing using transesophageal echocardiography (TEE) with multislice CT. As a result, the patients received a transcatheter heart valve that was too small.

The study was published in the June issue of the Journal of the American College of Cardiology: Cardiovascular Interventions.

Darren Mylotte, MB, MD, of McGill University in Montreal, and colleagues assessed adherence to oversizing principals in high-risk patients with severe aortic stenosis being treated with the self-expanding CoreValve system (Medtronic) at the German Heart Center in Munich. Valves require oversizing to ensure proper anchoring and sealing in a TAVR procedure. If a valve is too small, it may lead to paravalvular leak.

They evaluated 159 patients who underwent screening by CT before a TAVR procedure between Jan. 6, 2009, and June 6, 2010. TAVR sizing had been determined by TEE based on the standard sizing criteria. They retrospectively calculated diameters for each patient and applied the TEE-based size to recalculate the CT-derived oversizing. The primary outcome was paravalvular leak, defined as post-procedural aortic regurgitation of two or more or the need for post-implant balloon dilation.

Compared with CT-derived diameters, the TEE-derived parameters were smaller. TEE measurements considered 95.5 percent of the patients to be suitable for the 26mm or 29mm CoreValve, 10.8 percent for the 23mm valve and 10.8 percent for the 31mm valve. CT found 18 percent fewer patients suitable for the 26mm or 29mm valves and 11 percent more suitable for the 31mm valve. In addition, CT found 8.9 percent unsuitable for any of the sizes available at that time.

Mylotte et al calculated that only 51 percent of patients had the recommended valve oversizing. CT data showed significantly less valve oversizing in patients with paravalvular leak compared with patients without paravalvular leak. Adhering to the CT-based but not the TEE-based oversizing was a predictor of reduced paravalvular leak.

“Pre-procedural anatomical screening is of considerable importance for TAVR,” they wrote. “In particular, appropriate THV [transcatheter heart valve] sizing is recognized to be a key factor for optimizing patient outcomes.”

They argued that CT provides a more accurate annulus measurement compared with TEE, but they added their analysis did not look at 3D TEE, which may be on par with CT. They also pointed out that their study was retrospective and applied only to the CoreValve patients.