Sonicated X-ray contrast agents for quantitative myocardial contrast echocardiography--a critical approach

Contrast echocardiography with sonicated radiographic contrast agents has been used for the qualitative and quantitative determination of myocardial blood flow. One major problem has been the size of the microbubbles since only bubbles smaller than 8 microns are expected to pass the capillary bed and larger bubbles may obstruct the capillaries and, thus, alter myocardial blood flow. These techniques have been used for several years, but their reliability has not yet been assessed accurately. Five different methods for the production of sonicated radiographic contrast agents (methods 1-3 from the literature, and 4 and 5 from our laboratory; M1-5) were evaluated for their use in quantitative contrast echocardiography. The sonication of non-ionic X-ray contrast media was performed with a standard titanium probe (20 kHz) for methods 1-4, with variation in the sonication time and the number of sonication jets used for each method. In M5, we used bubbles that were produced by the insufflation of oxygen in the X-ray contrast agent; large (> 8 microns) bubbles were destroyed by sonication at 380 kHz (resonance method). Mean bubble size was determined by computerized videomicroscopy. The effect of bubble size on the backscatter of the ultrasonic signal was calculated for each method. Mean bubble size (+/- 1 SD) ranged between 11.5 +/- 4 microns and 16.1 +/- 14 microns for M1-M5. The best values, i.e., the smallest bubbles, were found with M4 (prepressurized contrast medium). Assuming capillary passage for bubbles smaller than 8 microns, only 14%-48% of the bubbles were smaller than 8 microns (M1-M5). The best results with regard to bubble size (< or = 8 microns) were observed with M5 (48% < or = 8 microns). In regard to the influence of bubble size on the backscatter of the ultrasonic signal, 56%-98.5% of the signal was produced by bubbles larger than 15 microns (M1-5) but the best results were obtained with M4. It is concluded that capillary-passage of sonicated microbubbles (< or = 8 microns) can be expected in only 14%-48% of the bubbles for the five different sonication techniques. More than 50% of all microbubbles produced by these techniques are larger than the expected 8 microns. These large bubbles are responsible for the backscatter of the ultrasonic signal in the vast majority of cases. Thus, the sonication of radiographic contrast agents appears to be inappropriate for the production of uniformly small microbubbles and, thus, this method is not suitable for quantitative measurements of coronary blood flow.