Israel Medical Association Journal

Cardiac tumors are rare and the majority are from a primary source outside of the heart. Most are found, incidentally, with echocardiography but often additional cardiac imaging is needed to refine the differential diagnosis. For this purpose, cardiac magnetic resonance imaging (MRI) and to a lesser extent cardiac computed tomography (CT) or ¹⁸F-fluorodeoxyglucose positron-emission tomography/computed tomography (¹⁸F-FDG PET/CT) are useful imaging modalities to better characterize a cardiac tumor and determine the likelihood of a neoplastic versus non-neoplastic origin. Cardiac CT may be useful to evaluate the effect of treatment while using ¹⁸F-FDG PET/CT to evaluate cardiac masses is under-studied but may be useful in patients who are already having a scan performed for oncologic reasons. It is through understanding the clinical context of a newly discovered cardiac mass, knowledge of the typical locations of various cardiac tumor types, combined with imaging techniques that avoid ionizing radiation that yield the greatest confidence in the noninvasive diagnosis of a cardiac mass

Background: Correct diagnosis of cardiac masses is a challenge in clinical practice. Accurate identification and differentiation between cardiac thrombi and tumors is crucial because prognosis and appropriate clinical management vary substantially.

Objectives: To evaluate the diagnostic performances of cardiac magnetic resonance imaging (CMR) in differentiating between cardiac thrombi and tumors.

Methods: A retrospective review of a prospectively maintained database of all CMR scans was performed to distinguish between cardiac thrombi and tumors during a 10 year period in a single academic referral center (2004–2013). Cases with an available standard of reference for a definite diagnosis were included. Correlation of CMR differentiation between thrombi and tumors with an available standard of reference was performed. Sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy were reported.

Results: In this study, 101 consecutive patients underwent CMR for suspicious cardiac masses documented on transthoracic or transesophageal echocardiography. CMR did not detect any cardiac pathology in 17% (17/101), including detection of anatomical variants and benign findings in 18% (15/84). Of the remaining 69 patients, CMR diagnosis was correlated with histopathologic result in 74% (51/69), imaging follow-up in 22% (15/69), and a definite CMR diagnosis (lipoma) in 4% (3/69). For tumors, diagnostic accuracy, sensitivity, specificity, PPV, and NPV were 96.6%, 98%, 86.6%, 96.2%, and 96.6%, respectively. For thrombi, the results were 93.6%, 86.7%, 98.04%, 92.9%, and 97%, respectively.

Conclusions: CMR is highly accurate in differentiating cardiac thrombi from tumors and should be included in the routine evaluation of cardiac masses.