Temporal change in PET myocardial flow reserve is predictive of cardiovascular outcomes

Abstract Introduction The association between serial measurements of myocardial flow reserve (MFR) and outcomes has received limited study. Purpose We sought to describe the trends of coronary blood flows over time in patients undergoing evaluation for coronary artery disease, as well as their association with cardiovascular outcomes. Methods We analyzed data from 948 positron emission tomography (PET) stress tests performed on 474 patients without prior history of heart transplant for the evaluation of coronary artery disease at our Hospital between 2016-2022. Patients were assigned to categories according to MFR trajectory based on previously established cut-off of 2.0 (low to low, low to high, high to low, high to high). Continuous variables are expressed as median interquartile range and rates as percentages. Long-term outcomes defined as death or myocardial infarction (MACE) were analyzed with the Kaplan-Meier method and Cox regression. C-statistic and net reclassification were used to assess model performance. Results Median interval time between tests was 804 days. Most common indications for the first and second PET were chest pain (68 and 65%), dyspnea (11 and 15%) and perioperative risk stratification (11 and 9%) without significant difference (P=0.33). MFR was similar in serial exams (2.1 1.7-2.6 vs 2.1 1.7,2.5, P=0.68, but rest (1 0.8, 1.3 vs 0.9 0.7, 1 P0.01) and stress flows (2.1 1.6, 2.8 vs 1.8 1.4, 2.2, P0.01) were both reduced on the second PET. In patients with revascularization between tests, MFR was similar in serial exams (1.6 1.3-2.0 vs 1.7 1.2,2.2, P=0.55), as was in patients without revascularization (2.2 1.8, 2.6 vs 2.1 1.8, 2.5, P=0.36). During the median follow-up time of 457 202,853 days, rate of MACE was 28% (51 events). At 4 years after the second PET, patients in the low to low and low to high group experienced worse outcomes. (Figure 1A). In Cox-regression multivariable analysis, there was an increase in MACE in patients in the low to low but not in the low to high group (Figure 1B). There was a decrease on the c-statistic for the above model when MFR trajectory was substituted with the baseline MFR value (0.76 to 0.67), indicating increased prognostic ability when serial measurements are taken into account. The model containing MFR trajectory correctly reclassifies 2.4% of patients that experience the outcome from low risk to high risk while down-classifying 0.06% of patients that do not experience the outcome from high to low risk, with a net reclassification index of 0.025. After rate pressure product correction of both rest and stress flow, there was no association between corrected MFR trajectory and outcomes when using a corrected MFR cut-off of 1.5 or 2. An increase in stress flow was not associated with better outcomes. Conclusion Serial MFR measurement may provide prognostic value in patients with low MFR on a PET stress test.