Assessment of the arterial input function for estimation of coronary flow reserve by single photon emission computed tomography: comparison of two different approaches

Attempts to estimate coronary flow reserve (CFR) with single photon emission computed tomography (SPECT) tracers have been recently made. We compared two different methods for the estimation of CFR by SPECT imaging. Fourteen patients with coronary artery disease underwent dipyridamole (99m)Tc-sestamibi SPECT and intracoronary Doppler within 5 days. Myocardial blood flow (MBF) was estimated by measurement of first transit counts in the right pulmonary artery (PA) and left ventricular (LV) chamber, and myocardial counts from SPECT images. Estimated CFR was expressed as the ratio of stress MBF to rest MBF. Rest and stress MBF obtained using first transit counts from PA were higher compared to that from LV chamber (rest: 1.05 +/- 0.38 vs 0.87 +/- 0.34 counts/pixel per s, respectively, p < 0.01 and stress: 1.34 +/- 0.45 vs 0.91 +/- 0.20 counts/pixel per s, respectively, p < 0.05). In the study vessels, CFR by Doppler was 1.39 +/- 0.42, and SPECT CFR obtained using first transit counts from PA and LV chamber were 1.36 +/- 0.43 and 1.16 +/- 0.39, respectively (p across categories NS). A significant relationship between SPECT CFR obtained using first transit counts from PA and CFR by Doppler was found (r = 0.85, p < 0.001). No relationship between SPECT CFR obtained using first transit counts from LV chamber and CFR by intracoronary Doppler was observed (r = 0.43, p = NS). SPECT-estimated CFR obtained using first transit counts from right PA is more accurate and correlates better with the results of intracoronary Doppler than estimated CFR obtained using arterial input function from LV chamber.