Long-term prognostic value of inducible and resting perfusion defects detected by single-photon emission computed tomography in the era of wide availability of coronary revascularization



Dominika Filipiak-Strzecka, Department of Cardiology, Medical University of Lodz, Kniaziewicza 1/5, 91-347 Lodz, Poland

E-mail: dominika.filipiak@gmail.com



To assess the long-term prognostic value of various types of perfusion defects detected by single-photon emission computed tomography (SPECT) in patients with stable angina.


Seven hundred and thirty two patients (299 men, mean age 57 ± 9 years) with suspected or known stable coronary artery disease underwent rest/exercise SPECT protocol using 99mTc-methoxyisobutylisonitrile (MIBI). All patients completed clinical follow-up regarding cardiac events (cardiac death, sudden cardiac death, acute coronary syndrome, revascularization, cardiac hospitalization) for a mean period of 58 ± 8 months. Event rates were analysed in subgroups defined according to the presence of fixed or inducible myocardial perfusion defects.


During the follow-up, 15 cardiac deaths were recorded, 13 of which were qualified as a sudden cardiac death, 59 acute coronary syndromes, 65 revascularizations and 209 cardiac hospitalizations. The presence of inducible perfusion defects on SPECT was associated with a high risk for occurrence of all analysed end points: sudden cardiac death (HR = 3·96, P = 0·01), cardiac hospitalization (HR = 1·5, P = 0·004), coronary syndrome (HR = 2·33, P = 0·001) and coronary revascularization (HR = 2·76, P = 0·0002), except for the cardiac death (P = 0·27). Resting perfusion defects were highly predictive for cardiac death (HR = 7·45; P = 0·001), but not for other cardiac events (= NS). The presence of any (inducible or resting) perfusion defects was associated with a high risk of all cardiovascular complications.


In long-term follow-up, SPECT proved highly predictive of cardiac events in patients with suspected or known CAD. In the revascularization era, cardiac death is most accurately related to the presence of resting perfusion defects, but all other cardiac events were better predicted by inducible perfusion defects.