Patient Population

Among the 423 patients with MS followed at our institution, we enrolled 101 patients according to the following inclusion criteria: mild to moderate MS of rheumatic origin, sinus rhythm, recent (<2 months) negative results on exercise treadmill ECG at peak heart rate, and no need for medical therapy. According to American guidelines, MS was considered mild when the mean gradient was <5 mm Hg and valve area was >1.5 cm ² and moderate when mean gradient was between 5 and 10 mm Hg and valve area was between 1 and 1.5 cm ² .

Exclusion criteria were New York Heart Association class ≥II, diabetes mellitus, blood arterial hypertension, dyslipidemia, coronary artery disease, arrhythmias (according to history, ECG at inclusion, 24-hour Holter recording, and exercise treadmill ECG), associated valvular lesions such as more than mild aortic or mitral insufficiency, aortic stenosis, or more than mild tricuspid regurgitation, pulmonary arterial hypertension (≥45 mm Hg), left ventricular (LV) and right ventricular systolic dysfunction (LV ejection fraction [EF] <55%, right ventricular systolic annular velocity <11.5 cm/sec), LV hypertrophy (both septal and posterior wall thicknesses >11 mm), lung disease, and inadequate echocardiograms. One hundred one patients met these criteria and were included in the study.

As a reference group, we studied 70 healthy adult subjects, comparable for age and sex, without detectable cardiovascular disease and/or risk factors (family history of cardiac diseases, arterial hypertension, diabetes mellitus, hypercholesterolemia, cigarette smoking, and previous cerebrovascular disease) and without cardiovascular structural or functional abnormalities (normal results on clinical examination, ECG, and echocardiography) who were not receiving any medication.

Echocardiography

2D Strain

Two-dimensional ϵ uses grayscale (B-mode) sector images and is based on frame-by-frame tracking of small rectangular image blocks with stable speckle patterns. A minimum frame rate of 50 Hz was required for reliable operation of this program, and frame rates of 50 to 90 Hz were used for routine grayscale imaging. The apical four-chamber view was obtained using the same ultrasound system and probe used for standard echocardiography. By tracing the endocardial contour on an end-systolic cavitary frame after defining the thickness of the region to be considered, the software will automatically track the atrial wall on subsequent frames. Adequate tracking can be real-time verified and corrected by adjusting the region of interest or the contour (increased or decreased the width for thicker or thinner walls, respectively).

Using this technique, analysis was performed from the four-chamber and two-chamber apical views for the segments of LA septum, LA lateral wall, LA inferior wall, LA anterior wall, and near LA roof, as considered for Doppler ϵ imaging in previous studies, and then were averaged.

Accordingly, for each segment, during LV systole, LA wall longitudinal lengthening was represented with a peak positive ϵ value; it is observed because of filling of the atrial cavity by blood coming from the pulmonary veins while the mitral valve is closed. Myocardial longitudinal shortening was represented with a peak negative ϵ value; it occurs during LV diastole, when the atrium empties into the LV cavity.

The tracking process and conversion to Lagrangian ϵ was performed offline using dedicated software (EchoPAC; GE Vingmed Ultrasound AS).

Although this software was validated for the measurement of LV ϵ, it has been previously used for measuring LA ϵ with high feasibility and good agreement in many studies.

Follow-Up

Serial prospective follow-up (4 years) was obtained in all patients by means of a physician-directed telephone interview using a standardized questionnaire. The physicians were blinded to patient echocardiographic results. Hospitalization for cardiac cause was confirmed by review of hospital medical record. All patients were interviewed at least every 3 months during the follow-up period. After inclusion, all patients were seen every 6 months at an outpatient clinic and at any time they reported symptoms. At each examination, standard 12-lead ECG was used, and an inquiry was made about symptoms. A 24-hour Holter recording was made every month and at any time the patient had any symptoms suggesting AF.

The end point of the study was the occurrence of AF on standard 12-lead ECG or on 24-hour Holter recording.

Statistical Analysis

All analyses were performed using SPSS version 11.0 (SPSS, Inc., Chicago, IL). To compare patients with MS versus referents, or patients with MS with events versus those without events, unpaired Student’s t tests were used for continuous variables, while for categorical variables, Fisher’s exact tests were used. Quantitative values are presented as mean ± SD.

The relations between quantitative variables were studied using linear regression. In addition, a multivariate analysis was performed using forward stepwise Cox regression analysis using the likelihood ratio test, with a P value of .10 as the threshold for entering or removing variables. The multivariate analysis was elaborated from variables identified as statistically significant by univariate analysis and those considered clinically relevant.

In the search for a diagnostic cutoff value for the occurrence of events at follow-up, receiver operating characteristic curve analysis was performed, and the area under the curve is reported, which is representative of the discriminatory ability of the variable cutoff. Sensitivity and specificity and the positive and negative predictive values of the best cutoff variable were calculated.

P values < .05 were considered statistically significant.

The cumulative probability of freedom from AF was calculated using Kaplan-Meier life-table analysis and compared between groups using the log-rank test.

Interobserver variability, expressed as the coefficient of variation, was assessed by analyzing four longitudinal regions in 25 different, randomly chosen subjects by two independent investigators. For intraobserver variability, four longitudinal regions were analyzed by one investigator two times within an interval of 4 weeks. The second round of intraobserver measures was blinded to results from initial measures.