Comparison of 3 tracer kinetic models for myocardial perfusion quantification by dynamic 13N-ammonia PET imaging

Introduction: One-compartment model and two-compartment model are two commonly used models to analyze time-activity curves for measurement of myocardial blood flow (MBF) by dynamic ¹³N-ammonia positron emission tomography (PET). We aimed to compare 3 PMod models (de Grado Model, 2 Compartments Model and UCLA 2 Parameters Model) for quantification of MBF and myocardial flow reserve (MFR).

Methods: We retrospectively enrolled 53 cases ( 45 coronary microvascular dysfunction patients, 3 obstructive coronary artery disease patients and 5 patients treated with percutaneous coronary intervention). All patients underwent dynamic ¹³N-ammonia PET imaging at rest and during vasodilation by adenosine. Rest MBF, stress MBF and MFR were calculated using 3 PMod models respectively.

Results: There were significant differences among 3 tracer kinetic models for calculating rest MBF, stress MBF and MFR (all P<0.05). For rest MBF, the value of 2 Compartments Model highest [ 1.27( 0.39)] and there was no significant difference between the other two models (P>0.05). For stress MBF, the value of de Grado Model lowest [ 2.41( 0.92)] and there was no significant difference between the other two models (P>0.05). For MFR, the value of UCLA 2 Parameters Model highest [ 2.90( 1.07)] and there was no significant difference between the other two models (P>0.05). However, the correlation was good among the 3 models for calculation of rest MBF (ρ from 0.541 to 0.878, all P< 0.05), stress MBF (ρ from 0.762 to 0.918, all P< 0.05) , and MFR (ρ from 0.668 to 0.913, all P< 0.05). Additionally, the agreement among 3 tracer kinetic models for calculating rest MBF, stress MBF and MFR was not good.

Conclusions: The 3 tracer kinetic models of PMod can not be used interchangeably for myocardial perfusion quantification by dynamic ¹³N-ammonia PET imaging.