Abstract
24120
Introduction: Myocardial perfusion imaging assesses severity of coronary artery disease (CAD) to guide management decisions. HeartSee PET cardiac software uses coronary flow capacity (CFC), created from pixel stress ml/min/g and coronary flow reserve, to predict patient 10-year survival and its improvement after revascularization (revasc).
Methods: This study analyzed a subset of the CENTURY randomized trial (NCT00756379) of comprehensive lifestyle modification, lipid-lowering drugs to target levels, and quantitative rest-stress PET myocardial perfusion to qualify severity of CAD, guide revascularization and track treatment response. 751 participants completed baseline and 60-month protocol PETs with 36 receiving revascularization after baseline. PET 10-year survival probabilities, observed and virtual revasc are used to assess the model’s accuracy for the revasc and non-revasc groups.
Results: There was no significant difference between standard (22/36-61%) vs comprehensive arms (14/36-39%) (p=0.14) or demographics except for slightly more diabetes (p=0.04). 32 baseline revascularizations were by PCI (89%) and 4 were by CABG (11%). Revasc vs non-revasc groups had more coronary calcium (36/36 (100%) vs (639/715 (89%) (p=0.041) and more prior revasc (23/36 (64%) vs no-revasc (233/715-33%) (p<0.001). Revasc also had a larger percent of left ventricle relative stress defects, CFC severe, moderate, and mild abnormalities and significant smaller normal territories (p=0.001, <0.001, 0.002, and <0.001). Baseline PET-derived 10-year survival probability was 85%±12% for the non-revasc group and 72%±18% for revasc group (p<0.001).
60-month follow-up between standard and comprehensive groups’ demographics and PET metrics remained constant. Despite 36 receiving intervention, their relative perfusion defect remained greater than the non-revasc group (4%±7% vs 1%±5% (p=0.012). PET-CFC maps improved in revasc group, but not in the non-revasc group (Figure 1). CFC severe reduced from a mean of 9%±14% to 2%±5% after revasc that is now comparable to the non-revasc group, 1%±3% (p=0.088), but remains significantly worse for moderate, mild, and normal (p=0.044, 0.003, and <0.001, respectively) (Figure 1). The cumulative distribution chart in Figure 1 summarizes the improvement of PET-CFC for the revasc group and stability of the non-revasc and CENTURY study arms.
10-year survival probability improved in the revasc group from 72%±18% to 78%±16% (p=0.031), but remains worse on average than the non-revasc group’s stable 85%±12% (p=0.011), likely due to incomplete revascularization of multivessel or diffuse CAD.
PET virtual revasc accurately predicted improvement of observed 10-year survival probability from 72% to 78% compared to observed survival probability after actual revascularization of 78% (p=0.92). Non-revasc group also maintained their survival probability throughout the 60-month trial with medical management (p=0.316).
Figure 2 illustrates a baseline and post-revasc 60-month follow-up PET-CFC participant comparison. While the severe stress-induced perfusion defect is markedly reduced by PCI, a small, severe basal-septal area of perfusion defect remains in the distribution of a "jailed" first septal perforator with moderate diffusely reduced average maximum absolute perfusion and absolute coronary flow reserve with myocardial steal with collaterals.
Conclusions: For HeartSee CFC, the virtual revascularization model predicted observed 10-year survival probability after actual revascularization as seen on 60-month PET. Providing a basis for guiding revascularization for single-focal, multi-stenosis, or diffuse coronary artery disease.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.