Application of indicator dilution principles to regional assessment of coronary flow reserve from digital angiography

Summary

Although cineangiography is widely utilized to assess the severity of coronary artery disease, conventional arteriography exhibits intra- and interobserver variability and provides no information regarding flow characteristics. We have investigated application of indicator-dilution principles to computer analysis of digital subtraction angiography (DSA) as a means to calculate coronary flow reserve (CFR). Using a background-corrected epicardial coronary region of interest, CFR is directly proportional to the ratio of the quantity of contrast administered under hyperemic and basal conditions, and inversely proportional to the area under the contrast time curve. Utilizing this indicator-dilution algorithm, normal canine coronaries exhibit a CFR ratio exceeding 4:1. In the presence of artificially induced stenoses, we compared the indicator-dilution method to surgically implanted electromagnetic flow (EMF) probes and found a good correlation between DSA and EMF, CFR, r = 0.86.

Initial studies utilized a single frame per cardiac cycle which provided inadequate temporal resolution for slow heart rates. The method was subsequently modified to utilize 30 frames per second ECG-gated DSA and hand injection of contrast. A separate validation study showed close correlation in the animal model between CFR measured by DSA and EMF probe, r = 0.86. We subsequently investigated this method in 21 patients (35 vessels). Normal CFR in seven patients without stenoses was 4.8:1 ± 0.65. Patients with unstable angina had much lower flow reserve averaging 1.7 ± 0.3. Angioplasty generally improved CFR. We have recently applied the indicator-dilution method to calculation of absolute coronary blood flow (CBF), in an animal model. In this application, we used paired injections of contrast, one in the left ventricle (30 to 40 ml) and a second subselectively in the coronary. By comparing the area under these two epicardial contrast intensity curves, the proportion of cardiac output perfusing the coronary could be calculated. CBF was calculated by measuring total cardiac output using a thermodilution catheter. A close correlation between EMF and DSA was confirmed, r = 0.89.

These data indicate that an indicator-dilution model can be applied to contrast injection in the coronary arteries and CFR or CBF can be calculated accurately in both an animal model and in a patient population. Further investigations will be required to determine the clinical utility of these findings.