Background: On the basis of the close association of left atrial compliance with pulmonary hypertension in patients with mitral stenosis, we hypothesized that: (1) Doppler-derived net atrioventricular compliance (C(n)) can be used for predicting the occurrence of mitral valve (MV) replacement or percutaneous mitral commissurotomy (PMC); and (2) determinants of exercise capacity are variable depending on C(n).
Methods: We consecutively enrolled 26 patients (22 women; mean age, 47.0 +/- 6.1 years; range, 35-59 years) with pure moderate or severe mitral stenosis, in whom comprehensive echocardiography was performed at rest. C(n) was derived from the equation that has been previously validated (ie, C(n) = 1270 x [MV area by 2-dimensional planimetry/mitral E-wave downslope]. Measurements of stroke volume and systolic pulmonary artery pressure were repeated immediately after symptom-limited treadmill exercise test within 60 seconds.
Results: On the basis of earlier reports, C(n) less than 4 mL/mm Hg was selected to define abnormally small C(n). During a mean duration of 24-month follow-up, C(n) of less than 4 mL/mm Hg could reliably predict the occurrence of either MV replacement or PMC, which was confirmed by multivariate logistic regression analysis. In patients with C(n) less than 4 mL/mm Hg, exercise duration was found to be closely correlated primarily with resting systolic pulmonary artery pressure (r = -0.73, P = .03), whereas in patients with C(n) greater than or equal to 4 mL/mm Hg, postexercise stroke volume was the only determinant of exercise capacity (r = 0.49, P = .04). When combining C(n) less than 4 mL/mm Hg with MV area less than 1.0 cm(2) as a cut-off value for predicting MV replacement or PMC, the sensitivity improved from 63.6% to 81.8% (P = .05).
Conclusions: Left atrial compliance estimated by C(n) with Doppler echocardiography allows us to effectively predict the future occurrence of MV replacement or PMC, with a special usefulness in patients with moderate mitral stenosis. In addition, determinants of exercise capacity were variable depending on the degree of C(n).