Patient-prosthesis mismatch is negligible with modern small-size aortic valve prostheses

Background. Concern has been raised about residual significant gradients wh en small aortic prostheses are used, particularly in patients with large bo dy surface areas. We studied the performance of six types of small aortic p rostheses using dobutamine stress echocardiography. Methods. Sixty-three patients (mean age, 67 +/- 7 years) who had undergone aortic valve replacement 17 +/- 6 months previously were studied. Two bilea flet mechanical prostheses (St. Jude Medical and CarboMedics: sizes, 19 mm and 21 mm) and two biological prostheses (Medtronic Intact and St. Jude Bio Implant: size, 21 mm) were evaluated. A graded infusion of dobutamine was g iven and Doppler studies of valve performance were carried out. Results. All prostheses except one biological valve had acceptable hemodyna mic performance under stress. Using regression modeling, gradient at rest w as the only variable found to predict gradient under stress (p < 0.001). Mo reover, the most important predictor of gradient at rest was valve design, which accounted for 72% of the variance (p < 0.001). This relationship was independent of valve size (19 mm or 21 mm) or material tie, mechanical or b iological). Body surface area accounted for 4% of the variance in gradient only. Conclusions. The main predictor of transprosthetic gradient is the inherent characteristics of each particular prosthesis, with relatively insignifica nt contribution from variations in body surface area. Patient-prosthesis mi smatch is not a problem of clinical significance when certain modern valve prostheses are used. (C) 1999 by The Society of Thoracic Surgeons.