Purpose. To develop an optimal approach for evaluating 13N-ammonia positron emission tomography combined with computed tomography (PET/CT) data which will have the highest diagnostic accuracy in the detection of coronary stenosis ≥50%.

Material and methods. We examined 122 patients (102 with coronary artery disease (CAD), 22 – healthy people) who underwent dynamic and static stress-PET/CT and coronary angiography. Diagnostic performance of the following indicators was determined: quantitative – coronary flow reserve (CFR), semi-quantitative – reversible perfusion defect (SDS), as well as the proposed combined approach, including the use of both indicators. An analysis of 306 coronary arteries was performed in the same group of patients, of which 200 had significant (≥50%) stenosis.

Results. The diagnostic performance of PET/CT in the detection of CAD due to atherosclerotic lesion was assessed: for CFR sensitivity was 69%, specificity – 100% and accuracy – 74%; for SDS – 79%, 70% and 78% respectively. At the same time the combined approach including the use of both indicators increases the diagnostic performance in the detection of CAD: sensitivity up to 83% and accuracy up to 86%, with maximum specificity – 100%.

The diagnostic performance in the detection of significant coronary stenosis (50-100%) was assessed: for CFR sensitivity was 65%, specificity – 80% and accuracy – 72%; for SDS – 56%, 81% and 67% respectively. The diagnostic performance of the combined approach was higher compared to standard methods: sensitivity was 78%, specificity – 80% and accuracy – 79%.

Conclusion. The combined approach for assessing CFR and SDS by PET/CT – quantitative and semi-quantitative analyzes – allowed us to significantly increase the diagnostic performance of the method both in the detection of CAD as well as a single coronary artery obstructive lesion. This became possible mainly due to the elimination of a number of disadvantages, which include doubtful cases of reduced 13N-ammonia accumulation in the posterolateral wall or the LV apex, as well as to overcome the “balanced ischemia” effect in the analysis of patients with multivascular CAD according to the standard semi-quantitative method.