Preoperative assessment of mitral valve abnormalities in left atrial myxoma patients using cardiac CT

BACKGROUND To retrospectively evaluate mitral valve abnormality in left atrial myxoma patients by using cardiac computed tomography (CT). MATERIAL AND METHODS Cardiac CT was performed in 56 patients with left atrial myxoma and 50 controls. Tumor and mitral valve characteristics were analyzed. The mitral valve parameters differences were compared between patients with myxoma and controls, myxoma with or without mitral valve obstruction, different obstruction degrees, respectively. Receiver operating characteristic analysis was performed to determine the cut-off values of abnormal mitral valve parameters for myxoma patients. Multiple linear regression, logistic regression models and cox regression analysis were used to determine factors associated with mitral valve abnormalities, mitral obstruction, mitral regurgitation and postoperative recovery, respectively. RESULTS Myxoma induced the dilation of mitral valve, with different results among different degrees of obstruction (p<0.001). Mitral valve parameters had relationship with myxoma parameters. The cut-off values for discriminating mitral valve abnormalities in myxoma patients were found. Some significant predictors for mitral obstruction were tumor pedicle-tumor volume and patient age (HR, 0.886-30.811; p = 0.011-0.043). Moreover, the predictor for mitral regurgitation was mitral annulus diameter in diastolic phase (HR, 20.862; 95%CI,1.331-327.100; p = 0.031). Some predictors associated with postoperative recovery of mitral regurgitation were age, mitral annulus area, mitral annulus diameter and mitral valve diameter cutoff value for diastolic phase (HR, 0.001-119.160; p = 0.012-0.028). CONCLUSION Cardiac CT is capable of quantitatively assessing myxoma characteristic and mitral valve abnormality induced by myxoma, thus providing guidance of operative management and postoperative evaluation.