RELIABILITY OF ECHOCARDIOGRAPHIC MEASUREMENTS OF MYOCARDIAL PERFUSIONUSING COMMERCIALLY PRODUCED SONICATED SERUM-ALBUMIN (ALBUNEX)

Objectives. The purpose of this study was to assess the reproducibilit y and pitfalls of intracoronary and aortic root sonicated albumin inje ctions, using time-intensity curves, in a large sample of normal dogs. Background. The utility of a new myocardial contrast echocardiographi c agent, sonicated serum albumin (Albunex), is currently under investi gation. However, the reproducibility, injection techniques and general pitfalls of this contrast agent have not been well characterized. Met hods. We administered sequential intracoronary and aortic root injecti ons (518 injections) of sonicated albumin in 25 closed chest normal do gs to measure the effectiveness and reproducibility of this product. T ime-intensity curves, as a measure of myocardial perfusion, were deriv ed and quantified using as on-line videodensitometric analysis system and two-dimensional echocardiography. Measurements included peak inten sity, area under the curve, half-time of descent, alpha-parameter and transit time within a 31- x 31-pixel ''region of interest'' in the ant erior septum. Analyses provided 80% power and a type I error protectio n of 95%. Results. The best reproducibility of the variables was half- time of descent for aortic root injections (coefficient of variation [ CV] 20%) and peak intensity for intracoronary injections (CV 25%), whe reas aortic root area under the curve showed the most variability (CV 41%). Analysis of variance for repeated measures of serial intracorona ry and aortic root injections showed no significant systematic variabi lity within subjects for the measured variables. In a comparison betwe en intracoronary and aortic root injection sites, paired t tests showe d no significant difference for mean values between these two techniqu es. There was also no statistically significant difference between man ual versus power intracoronary injections. Finally, there was no signi ficant difference among three injection rates (1, 2 and 3 ml/s) in pai red intracoronary injections, nor was there a difference among injecti on rates in paired aortic root injections, except for a lower peak int ensity with a 1-ml/s injection rate compared with a 2-ml/s injection r ate (p = 0.01). Potential pitfalls include preparation of sonicated al bumin, delivery techniques and measurement variables. Conclusions. We conclude that the results of serial injections of sonicated albumin sh ow no systematic change or trend in normal dogs. Both intracoronary an d aortic root injections at standard injection rates by hand or power injector can be used to quantify time-intensity curves, as measure of myocardial perfusion, with similar variability ranging from 20% to 41% . These results are important in the human model, especially after cor onary interventions.