Evaluation of cardiac masses by real time perfusion imaging echocardiography

Author(s)

Eliza Kaori Uenishi

• Document type: Doctoral Thesis
• Press: São Paulo.
• Institution: Universidade de São Paulo (USP). Faculdade de Medicina (FM/SBD)
• Defense date: 2011-05-11

Examining board members

Wilson Mathias Júnior; Francisco Aparecido Belfort; Márcia Azevedo Caldas; Valdir Ambrósio Moises; Carlos Eduardo Rochitte

• Advisor: Wilson Mathias Júnior

Abstract

Background: Cardiac masses (CM) can be tumors, thrombi or pseudotumors. Evaluation of their vascularization might be an additional tool to perform a differential diagnosis. In the present study we demonstrated the diagnostic value of perfusion echocardiography for CM characterization, by qualitative and quantitative analyses of perfusion. Methods: We prospectively studied 107 patients, who were classified into 4 groups: 33 thrombus, 23 malignant tumors (MT), 24 benign tumors (BT) and 6 pseudotumors, of which 21 were excluded because no definitive diagnosis could be confirmed. Perfusion evaluation was performed by contrast echocardiography with real time perfusion imaging using microbubbles. A group of patients (32) was selected for a complementary study using dipyridamole to evaluate mass flow reserve. Qualitative and quantitative analyses were performed by two independent observers. Parameters for qualitative analysis were perfusion intensity (0-3 score), microvascular refilling velocity (0-2 score), central or peripheral perfusion pattern (0-2 score), and presence of areas of necrosis (0 or 1 score). The two parameters for quantification of masses were microvascular blood volume (A), and regional microvascular flow which is the product of blood flow velocity and vomume (A). Results: The most frequent pattern for the thrombi group in the qualitative analysis was absence of perfusion (81.9%), followed by no perfusion velocity (81.9%), and no areas of necrosis (93.4%), whilst among tumors there was predominance of discrete perfusion (62.3%), with slowed velocity (64.2%), and areas of necrosis (30.2%). Qualitative analysis, perfusion velocity showed intraobserver variability 20%, presence of areas of necrosis of 25% and perfusion pattern of 45%. In the quantitative analysis, the thrombi group was shown to have A and Ax values significantly smaller compared to the tumor group: Thrombi: A = 0.08 (0.01-0.22dB); Ax = 0.03 (0.010.14dB/s-1); MT: A = 2.78 (1.31-7.0dB); Ax = 2.0 (0.995.58dB/s-1); BT: A = 2.58 (1.24-4.55dB); Ax = 1.18 (0.453.4dB/s-1). When only the tumor groups with the use of dipyridamole were compared, MT was shown to have greater microvascular blood volume (A): A = 4.18 (2.14-7.93dB); Ax = 2.46(1.424.59dB/s-1), BT: A = 2.69 (1.11-4.265dB); Ax = 1.55 (0.555.50dB/s-1). Analysis of the ROC curve showed that an area of 0.95 for a microvascular blood volume of A < 0.65 dB predictive curve on perfusion echocardiography, both with and without dipyridamole, predicts thrombi, and so does a <0.30dB/s-1microvascular blood flow (Ax), area under curve = 0.94. In order to distinguish MT from BT, a >3.28dB microvascular blood volume (A) using dipyridamole was predictor of MT (area under curve = 0.75). Conclusion: The echocardiographic study to evaluate CM perfusion showed that qualitative analysis is reproducible diagnostic approach for diagnosing thrombi. Cardiac tumors show greater microvascular volume and regional blood flow when compared with thrombi. Dipyridamole quantitative stress mass perfusion was useful to differentiate MT from BT (AU)