Reproducibility of quantitative myocardial perfusion and coronary flow capacity by positron emission tomography: 3D digital silicon photomultiplier solid state vs. legacy 2D analogue systems for clinical practice and trials.

Abstract

Aims: Quantitative rest-stress myocardial perfusion in millilitres per minute per gram among multiple 2D and 3D positron emission tomography-computed tomography (PET-CT) scanners is essential for personalized cardiac management and clinical trials. Accordingly, this study reports the accuracy and precision of quantitative rest-stress millilitres per minute per gram and coronary flow capacity among 2D and two different digital 3D silicon photomultiplier (SiPM) PET-CT scanners for quantifying the severity of coronary pathophysiology for clinical trials or guiding interventions vs. medical treatment.

Methods and results: One hundred seventy-one participants underwent 748 paired serial rest or stress PET perfusion imaging in the same person on 'same day' or 'different days' using rubidium-82 (Rb-82) pharmacologic stress on 2D and two different digital 3D SiPM PET-CT scanners for global myocardial perfusion in millilitres per minute per gram. For methodological variability of 66 'same-day' serial paired PETs in the same person by 2D and two different 3D SiPM PET-CT scanners, rest-stress global myocardial millilitres per minute per gram had no significant bias (P = 0.464, mean difference 0.014 ± 0.21 mL/min/g) with coefficient of variation (COV) of ±14%. For methodological plus biological variability of 154 'different-day' serial paired PETs, rest-stress global perfusion had no significant bias (P = 0.136), mean difference (0.028 ± 0.33), and COV of ±20%. Coronary flow reserve had a small bias of 0.095 ± 0.57 (P = 0.041) and COV of ±20%. Coronary flow capacity was not different by Kolmogorov-Smirnov test (P = 0.99).

Conclusion: For quantifying myocardial perfusion in the same person on 'same day' or 'different days' using Rb-82, 3D SiPM PET-CT is comparably reproducible to analogue 2D PET-CT with the HeartSee perfusion model as the basis for quantifying physiologic severity of coronary heart disease to guide clinical decision-making or randomized clinical trials confirming these outcomes.